effects of inlet conditions on diffuser outlet performance

UNLV Theses/Dissertations/Professional Papers/Capstones

5-1-2011

Effects of inlet conditions on diffuser outletperformanceZaccary A. PootsUniversity of Nevada, Las Vegas

Follow this and additional works at: http://digitalscholarship.unlv.edu/thesesdissertationsPart of the Acoustics, Dynamics, and Controls Commons, Electro-Mechanical Systems

Commons, and the Statistical Models Commons

This Thesis is brought to you for free and open access by Digital Scholarship@UNLV. It has been accepted for inclusion in UNLV Theses/Dissertations/Professional Papers/Capstones by an authorized administrator of Digital Scholarship@UNLV. For more information, please [email protected].

Repository CitationPoots, Zaccary A., "Effects of inlet conditions on diffuser outlet performance" (2011). UNLV Theses/Dissertations/Professional Papers/Capstones. Paper 1431.

http://library.unlv.edu/?utm_source=digitalscholarship.unlv.edu%2Fthesesdissertations%2F1431&utm_medium=PDF&utm_campaign=PDFCoverPages

http://library.unlv.edu/?utm_source=digitalscholarship.unlv.edu%2Fthesesdissertations%2F1431&utm_medium=PDF&utm_campaign=PDFCoverPages

http://digitalscholarship.unlv.edu/thesesdissertations?utm_source=digitalscholarship.unlv.edu%2Fthesesdissertations%2F1431&utm_medium=PDF&utm_campaign=PDFCoverPages

http://digitalscholarship.unlv.edu/thesesdissertations?utm_source=digitalscholarship.unlv.edu%2Fthesesdissertations%2F1431&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/294?utm_source=digitalscholarship.unlv.edu%2Fthesesdissertations%2F1431&utm_medium=PDF&utm_campaign=PDFCoverPages




http://digitalscholarship.unlv.edu/thesesdissertations/1431?utm_source=digitalscholarship.unlv.edu%2Fthesesdissertations%2F1431&utm_medium=PDF&utm_campaign=PDFCoverPages

mailto:[email protected]

EFFECTS OF INLET CONDITIONS ON DIFFUSER

OUTLET PERFORMANCE

by

Zaccary A. Poots

Bachelor of Science in Mechanical Engineering University of Nevada, Las Vegas

2009

A thesis document submitted in partial fulfillment of the requirements for the

Master of Science in Mechanical Engineering Department of Mechanical Engineering

Howard R. Hughes College of Engineering

Graduate College University of Nevada, Las Vegas

May 2011

ii

THE GRADUATE COLLEGE We recommend the dissertation prepared under our supervision by Zaccary A. Poots entitled Effects of Inlet Conditions on Diffuser Outlet Performance be accepted in partial fulfillment of the requirements for the degree of Master of Science in Mechanical Engineering Douglas Reynolds, Ph. D., Committee Chair Brian Landsberger, Ph. D., Committee Member Samir Moujaes, Ph. D., Committee Member Sandra Catlin, Ph. D., Graduate Faculty Representative Ronald Smith, Ph. D., Vice President for Research and Graduate Studies and Dean of the Graduate College May 2011

iii

ABSTRACT

Effects of Inlet Conditions on Diffuser Outlet Performance

by

Zaccary A. Poots

Dr. Douglas Reynolds, Examination Committee Chair Professor of Mechanical Engineering

University of Nevada, Las Vegas

Building air distribution terminal system designers and system installers require

accurate quantitative information on the performance of the installed system to achieve

optimum efficiency and levels of human comfort. This requires field installation

adjustment values from published ideal pressure loss, air distribution and sound

generation installation performance. This study documents the air output performance of

different installation configurations of six types of ceiling diffusers and compares the

results to performance when installed according to ANSI/ASHRAE Standard 70-2006. A

diffuser inlet supply plenum was designed for optimum flow and used to acquire a

baseline set of data covering the six types of diffusers at different inlet neck sizes and

inlet airflow rates. Full scale laboratory testing of typical field installation variations was

completed for the same conditions with variations in damper installation, duct approach

angle, duct type, duct vertical height above the diffuser and duct branch to main supply

duct installation. A set of simple algorithms were developed that can be used to easily

predict how an inlet configuration would affect the performance of a wide variety of

installation conditions.

iv

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my primary advising professor, Dr. Brian J.

Landsberger, for the wisdom he bestowed upon me, the constant explanations that

seemed to make things much more approachable and his desire to help me succeed.

Without his constant guidance, his knowledge and his friendship, I would have never

managed to get this far. I wish you nothing but the best in your future endeavors.

I would also like to thank Dr. Douglas D. Reynolds and the rest of my thesis

committee members for their contributions and guidance in my research. Thank you to

the entire staff of the mechanical engineering department at UNLV and to the folks at the

Las Vegas Valley Water District, with whom I interned, for their time and patience. To

my fellow peers, without whom I would not have enjoyed my time at UNLV as much as I

did, thank you for making my college experience something I will never forget.

Finally, I would like to extend my deepest gratitude to my family and friends for their

love and support. If not for them, my pursuit to further my education may have ended

sooner rather than later. Most of all, I would like to thank my parents, Lisa and Gregg

Poots, for everything they have done for me throughout my life and college career. I

could never thank you enough for helping make me the man I am today.

v

TABLE OF CONTENTS

ABSTRACT....................................................................................................................... iii

ACKNOWLEDGEMENTS............................................................................................... iv

LIST OF FIGURES .......................................................................................................... vii

LIST OF TABLES.............................................................................................................. x

CHAPTER 1 INTRODUCTION .................................................................................... 1

CHAPTER 2 DIFFUSER SUPPLY PLENUM OPTIMIZATION ................................ 5 2.1 Background ........................................................................................................... 5 2.2 Objective ............................................................................................................... 6 2.3 Methodology ......................................................................................................... 7 2.4 Results................................................................................................................. 16 2.5 Analysis............................................................................................................... 20

CHAPTER 3 BASELINE DIFFUSER CHARACTERIZATION................................ 24 3.1 Background ......................................................................................................... 24 3.2 Objective ............................................................................................................. 25 3.3 Methodology ....................................................................................................... 25 3.4 Results................................................................................................................. 35 3.5 Analysis............................................................................................................... 43

CHAPTER 4 FIELD CONDITION AIR OUTLET PERFORMANCE....................... 52 4.1 Background ......................................................................................................... 52 4.2 Objective ............................................................................................................. 53 4.3 Methodology ....................................................................................................... 54 4.4 Results................................................................................................................. 60 4.5 Analysis............................................................................................................... 66

CHAPTER 5 CLOSE COUPLING AIR OUTLET PERFORMANCE ....................... 88 5.1 Background ......................................................................................................... 89 5.2 Objective ............................................................................................................. 89 5.3 Methodology ....................................................................................................... 89 5.4 Results................................................................................................................. 96 5.5 Analysis............................................................................................................. 103

CHAPTER 6 CONCLUSIONS AND RECOMMENDATIONS............................... 115 6.1 Diffuser Supply Plenum Design ....................................................................... 115 6.2 Standard 70 Characterization ............................................................................ 116 6.3 Field Installations.............................................................................................. 117 6.4 Close Coupling.................................................................................................. 118 6.5 Recommendations for Designers and Installers................................................ 119

vi

APPENDIX A TEST FACILITY ............................................................................... 120

APPENDIX B FIELD INSTALLATION TEST RESULTS DIFFUSER DISCHARGE

AIRFLOW ZONE PLOTS................................................................. 136

APPENDIX C CLOSE COUPLING TEST RESULTS DIFFUSER DISCHARGE

AIRFLOW ZONE PLOTS................................................................. 159

BIBLIOGRAPHY........................................................................................................... 171

VITA............................................................................................................................... 173

vii

LIST OF FIGURES

Figure 1. Picture of the test plenum. ............................................................................. 10 Figure 2. Pictures of the different types (left) and the different sizes (right) of the flow

equalization disks. ......................................................................................... 11 Figure 3. Diffuser inlet configuration. .......................................................................... 12 Figure 4. Diffuser inlet variation testing configuration. ............................................... 13 Figure 5. Actual test configuration using TSI Ventilation Meter. ................................ 13 Figure 6. Signal to noise ratio results of the noise experiment. .................................... 15 Figure 7. Main effects plots for the optimization test. .................................................. 17 Figure 8. Interaction plots for the optimization test...................................................... 20 Figure 9. Duct height added from diffuser inlet to inlet cone....................................... 22 Figure 10. Standard 70 vertical ducted method. ............................................................. 27 Figure 11. Boom microphone configuration................................................................... 28 Figure 12. Array of draft meters. .................................................................................... 30 Figure 13. Standard 70 Zone Plot. .................................................................................. 34 Figure 14. Velocity profile of 12 inch square diffuser at 1200 fpm. .............................. 35 Figure 15. Velocity profile of 8 inch square diffuser at 1200 fpm. ................................ 36 Figure 16. Noise criterion level for 12 inch square diffuser at 1200 fpm....................... 37 Figure 17. Plenum total pressure for 12 inch square diffuser at 1200 fpm..................... 37 Figure 18. Standard 70 zone plot for all square diffusers. .............................................. 43 Figure 19. Standard 70 zone plot for plaque diffusers.................................................... 45 Figure 20. Standard 70 zone plot for perforated round diffusers.................................... 46 Figure 21. Standard 70 zone plot for perforated round/modular core diffusers. ............ 47 Figure 22. Standard 70 zone plot for round diffusers. .................................................... 48 Figure 23. Standard 70 zone plot for louvered diffusers. ............................................... 49 Figure 24. Cross flow scan pattern for the forward side of a diffuser. ........................... 50 Figure 25. Cross flow scan of 8” perforated round diffuser using

vertical ducted method. ................................................................................. 51 Figure 26. Experimental setup for field installation condition laboratory testing. ......... 56 Figure 27. Standard 70 throw data at 1200 fpm for 12” square diffuser. ....................... 60 Figure 28. Standard 70 zone plot for 12” square diffuser, where (x1,y1)

symbolizes start of zone 3 (yellow dot) and (pink dot) gives the x2 value when y2 is 0.2. .......................................................................... 62

Figure 29. Field installation run #3 forward throw at 1200 fpm. ................................... 63 Figure 30. Field installation run #3 backward throw at 1200 fpm.................................. 64 Figure 31. Zone plot of field install run #3 forward throw 1200 fpm. ........................... 64 Figure 32. Zone plot of field install run #3 backward throw 1200 fpm.......................... 65 Figure 33. Main effects plot for field installations forward throw. ................................ 67 Figure 34. Interaction plots for field installations forward throw................................... 68 Figure 35. Experimental setup for close coupling laboratory testing. ............................ 91 Figure 36. Top view of close coupling scan pattern. ...................................................... 93 Figure 37. Example close coupling configuration. ......................................................... 95 Figure 38. Standard 70 throw data at 1200 fpm for 12” square diffuser. ....................... 97 Figure 39. Close coupling forward throw data at 1150 fpm (in main duct) for Run #3. 97 Figure 40. Close coupling backward throw data at 1150 fpm (in main duct)

viii

for Run #3...................................................................................................... 98 Figure 41. Run #3 forward throw zone plot data. ........................................................... 99 Figure 42. Run #3 backward throw zone plot data. ...................................................... 100 Figure 43. Zone plot for forward throw of Run #1 without ceiling.............................. 101 Figure 44. Styrofoam ceiling for Run #1 redo.............................................................. 102 Figure 45. Zone plot for forward throw of Run #1 with ceiling. .................................. 102 Figure 46. Comparison between change in sound [dB] and main duct velocity. ......... 110 Figure 47. Side view of the UNLV throw room [15]. .................................................. 123 Figure 48. Labview main interface for monitoring system conditions. ........................ 124 Figure 49. Interface for monitoring throw room surface temperatures. ....................... 125 Figure 50. Throw room LabView motion controller interface. .................................... 125 Figure 51. Throw velocity sensor path setup interface. ................................................ 126 Figure 52. Calibrated nozzle chamber versus Ebtron unit. ........................................... 130 Figure 53. VAV unit calibration. .................................................................................. 131 Figure 54. Plenum pressure test. ................................................................................... 133 Figure 55. Published NC contour curves. ..................................................................... 134 Figure 56. Sound capture interface using the boom microphone. ................................ 135 Figure 57. Forward throw for Run #1........................................................................... 137 Figure 58. Backward throw for Run #1. ....................................................................... 138 Figure 59. Forward throw for Run #2........................................................................... 138 Figure 60. Backward throw for Run #2. ....................................................................... 139 Figure 61. Right side throw for Run #2. ....................................................................... 139 Figure 62. Forward throw for Run #3........................................................................... 140 Figure 63. Backward throw for Run #3. ....................................................................... 140 Figure 64. Forward throw for Run #4........................................................................... 141 Figure 65. Backward throw for Run #4. ....................................................................... 141 Figure 66. Right side throw for Run #4. ....................................................................... 142 Figure 67. Forward throw for Run #5........................................................................... 142 Figure 68. Backward throw for Run #5. ....................................................................... 143 Figure 69. Left side throw for Run #5. ......................................................................... 143 Figure 70. Forward throw for Run #6........................................................................... 144 Figure 71. Backward throw for Run #6. ....................................................................... 144 Figure 72. Forward throw for Run #7........................................................................... 145 Figure 73. Backward throw for Run #7. ....................................................................... 145 Figure 74. Forward throw for Run #8........................................................................... 146 Figure 75. Backward throw for Run #8. ....................................................................... 146 Figure 76. Forward throw for Run #9........................................................................... 147 Figure 77. Backward throw for Run #9. ....................................................................... 147 Figure 78. Right side throw for Run #9. ....................................................................... 148 Figure 79. Forward throw for Run #10......................................................................... 148 Figure 80. Backward throw for Run #10. ..................................................................... 149 Figure 81. Forward throw for Run #11......................................................................... 149 Figure 82. Backward throw for Run #11. ..................................................................... 150 Figure 83. Forward throw for Run #12......................................................................... 150 Figure 84. Backward throw for Run #12. ..................................................................... 151 Figure 85. Forward throw for Run #13......................................................................... 151

ix

Figure 86. Backward throw for Run #13. ..................................................................... 152 Figure 87. Forward throw for Run #14......................................................................... 152 Figure 88. Backward throw for Run #14. ..................................................................... 153 Figure 89. Right side throw for Run #14. ..................................................................... 153 Figure 90. Forward throw for Run #15......................................................................... 154 Figure 91. Backward throw for Run #15. ..................................................................... 154 Figure 92. Forward throw for Run #16......................................................................... 155 Figure 93. Backward throw for Run #16. ..................................................................... 155 Figure 94. Forward throw for Run #17......................................................................... 156 Figure 95. Backward throw for Run #17. ..................................................................... 156 Figure 96. Right side throw of Run #17. ...................................................................... 157 Figure 97. Left side flow for Run #17. ......................................................................... 157 Figure 98. Forward throw for Run #18......................................................................... 158 Figure 99. Backward throw for Run #18. ..................................................................... 158 Figure 100. Forward throw for Run #1 with no ceiling.................................................. 160 Figure 101. Backward throw for Run #1 with no ceiling. .............................................. 161 Figure 102. Forward throw for Run #1 with ceiling....................................................... 161 Figure 103. Backward throw for Run #1 with ceiling. ................................................... 162 Figure 104. Forward throw for Run #2........................................................................... 162 Figure 105. Backward throw for Run #2. ....................................................................... 163 Figure 106. Forward throw for Run #3........................................................................... 163 Figure 107. Backward throw for Run #3. ....................................................................... 164 Figure 108. Forward throw for Run #4........................................................................... 164 Figure 109. Backward throw for Run #4. ....................................................................... 165 Figure 110. Forward throw for Run #5........................................................................... 165 Figure 111. Backward throw for Run #5. ....................................................................... 166 Figure 112. Forward throw for Run #6........................................................................... 166 Figure 113. Backward throw for Run #6. ....................................................................... 167 Figure 114. Forward throw for Run #7........................................................................... 167 Figure 115. Backward throw for Run #7. ....................................................................... 168 Figure 116. Forward throw for Run #8........................................................................... 168 Figure 117. Backward throw for Run #8. ....................................................................... 169 Figure 118. Forward throw for Run #9........................................................................... 169 Figure 119. Backward throw for Run #9. ....................................................................... 170

x

LIST OF TABLES

Table 1. Test Parameter and Noise Condition List for Test Array. .............................. 10 Table 2. Test Array with One 4-Level Parameter, Four 3-Level Parameters And One

Noise Parameter at 2 Levels. .......................................................................... 16 Table 3. Analysis of variation for standard deviation for the optimization test. .......... 19 Table 4. Verification testing results. ............................................................................. 21 Table 5. Effect of added height variations for 12-inch diffuser inlet............................ 22 Table 6. 1/1 and 1/3 octave band center frequencies. ................................................... 29 Table 7. Vertical draft meter sensor locations. ............................................................. 31 Table 8. Test parameters and noise conditions with corresponding states. .................. 32 Table 9. Test array for Standard 70 testing with output measures................................ 33 Table 10. Final results for square diffuser at corresponding velocities. ......................... 38 Table 11. Final results for plaque diffuser at corresponding velocities. ......................... 38 Table 12. Final results for perforated round diffuser at corresponding velocities.......... 39 Table 13. Final results for perforated square/modular core diffuser at corresponding

velocities. ........................................................................................................ 40 Table 14. Final results for round diffuser at corresponding velocities. .......................... 41 Table 15. Final results for louvered diffuser at corresponding velocities....................... 42 Table 16. Plaque total pressure and sound comparisons of manufacturer data and

Standard 70 data from this experiment. .......................................................... 44 Table 17. Modular core total pressure and sound comparisons of manufacturer and

Standard 70 data.............................................................................................. 47 Table 18. Test parameters and noise conditions with different states. ........................... 58 Table 19. Test array for field condition variations testing.............................................. 59 Table 20. Output array at high state noise condition for the

field installation experiment. .......................................................................... 59 Table 21. Diffuser output throw ratios, noise criteria level changes and total pressure

ratios for field installations compared to Standard 70 data............................. 66 Table 22. Predicted values and corresponding errors for forward throw. ...................... 72 Table 23. Verification testing configurations. ................................................................ 73 Table 24. Results of verification testing configurations for forward throw. .................. 73 Table 25. Predicted and corresponding values for backward throw............................... 77 Table 26. Predicted and corresponding error values for pressure................................... 81 Table 27. Verification testing configurations. ................................................................ 81 Table 28. Results of verification testing for pressure. .................................................... 82 Table 29. Predicted values and corresponding errors for sound level. ........................... 85 Table 30. Verification testing configurations. ................................................................ 86 Table 31. Results of verification testing for sound level. ............................................... 86 Table 32. Summary of standard deviations for field-testing analysis............................. 87 Table 33. Test parameters and noise conditions with different states. ........................... 94 Table 34. Test array for close coupling experiment. ...................................................... 95 Table 35. Output array for the close coupling experiment. ............................................ 96 Table 36. Diffuser output throw ratios and noise criteria level changes for close coupling compared to Standard 70 data. .......................... 100 Table 37. Predicted values and corresponding error for forward throw. ...................... 105

xi

Table 38. Predicted values and corresponding errors for backward throw................... 107 Table 39. Predicted values and corresponding errors for total pressure ratio............... 109 Table 40. Adjusted NC levels that account for damper and not main duct velocity. ... 112 Table 41. Predicted values in dB and corresponding errors for sound (NC) level. ...... 114 Table 42. Throw room instrumentation used along with NI LabView......................... 127 Table 43. Testing equipment used in part with throw room. ........................................ 128 Table 44. Systems used in part with the throw room.................................................... 129 Table 45. Inlet conditions for field testing runs............................................................ 137 Table 46. Inlet conditions for close coupling runs........................................................ 160

1

CHAPTER 1

INTRODUCTION

It is widely known that under typical field conditions ceiling diffusers and other air

outlets are typically installed with inlet conditions significantly different from those

specified in Standard 70 resulting in performance differences from manufacturer

published performance data. Diffusers often have a flexible duct or a direct elbow duct

connection. The air duct running to the diffuser may have a hard radius or an angled

entrance into the device, and many have an air-balancing device at or near the diffuser

inlet. These inlet conditions can dramatically change the performance of outlets

compared to data obtained following Standard 70.

The primary information sources for VAV duct design are ASHRAE’s HVAC

Applications Handbook [1], California Energy Commission’s (CEC) Advanced Variable Air

Volume Design Guide [2], and the ASHRAE Handbook of Fundamentals. [3] Information

available covering the performance of the air distribution system section from the VAV box

to the diffuser includes duct design issues, performance issues, and installation problems. A

recent article in HVAC&R Research by Landsberger covers energy and acoustic performance

effects of installation variations. [4] Since the mid 1950’s, detailed measurements of

acceptable mean diffuser velocity speeds and air temperatures in occupied spaces have been

available to designers for ventilation spaces. [5]

Dynamics of air movement can have an effect on the air distribution system performance.

Supply air traveling within ductwork develops considerable momentum. When the supply-air

duct empties its air into the air-conditioned space, this momentum is utilized to help mix the

supply air with the room air. How air is delivered and removed from a space is known as

room air distribution. [6] This helps to ensure homogeneous or isothermal temperature and

2

air movement throughout the occupied zone of the conditioned space. This is accomplished

by understanding the manipulation of an outlet’s throw, drop and spread characteristics.

Throw of air is the horizontal and vertical distance a supply-air jet travels before reaching a

specified air velocity value after leaving the outlet. The drop of air is the vertical distance the

jet travels at its lower edge before reaching the end throw value. The spread of air indicates

the divergence of the air-stream after leaving the outlet. This knowledge leads into the

description of the flow as it propagates away from the diffuser into the four zones of

expansion. [7] These zones of expansion play a major roll in the analysis of the supplemental

data recorded during this project.

Installation variations that can result in significant performance variation from ASHRAE

Standard 70 installation predominantly concern the length and type of the duct branch, how

duct turns are accomplished and how the duct approaches the diffuser. Without sufficient

length to develop a uniform flow profile, the flow in duct branches too close to the VAV

terminal or in a previous branch being non-uniform, an increase in pressure loss is often the

result. If elbows and junctions, such as those made to avoid obstructions in the path of the

duct work, are not constructed with minimal friction effects, pressure loss and

aerodynamically generated noise will increase. [1] Round flexible ducts can serve as

transmitters of breakout sound and also be effective attenuators of upstream sound

sources. [1]

The duct approach to the diffuser is also very important, since detrimental effects of

improper duct approach cannot be corrected by the diffuser itself. Accepted guidance states

that the velocity of the air stream should be as uniform as possible over the entire outlet

connection to the duct and must be perpendicular to the outlet face. However, few outlets are

installed in this manner. [1,3] Flexible duct connections at the inlet of the diffuser can

3

increase pressure drop and non-uniform air distribution from the diffuser outlet. [1] An

elbow, transition or damper too close to the diffuser inlet, can result in highly directional

airflow, increased sound level and increased pressure drop across the outlet of standard

diffusers. [1]

Each diffuser manufacturer has published sound levels, for a given diffuser and flow rate,

under ideal conditions, per ASHRAE Standard 70. However, duct connections encountered in

the field result in significantly different and usually higher sound levels for the same airflow

rate. One can assume a 5 point increase in NC for typical field installations compared to

published data. [8] For example, an offset of the flexible duct connection between the

diffuser and the supply duct can increase the sound power level as much as 12 to 15 dB. [1]

A recent laboratory investigation found that a ninety degree elbow from a horizontal duct run

to the diffuser, with minimal vertical run, on average increases diffuser sound levels by 6 dB,

while a hard turn in the duct run, even several feet from the diffuser, added 4 dB to the sound

level. [4] Also, diffusers with a perforated face have been seen to have a higher sound

distortion with poor inlet conditions than diffusers with large open cones.

The magnitude of the sound and performance effects associated with installation

variations will also depend on variations out of the control of the installer such as

variable flow rates. Diffusers are designed to optimally distribute the air at some

particular load condition and air volume, but in a typical VAV type installation, air

volume rate will vary. Consequently, diffuser throw, room airflow velocity and sound

levels can be significantly different from that specified design point. [1]

This research, funded by ASHRAE, Inc. under project 1335-RP, identified

quantitative data on the performance differences between ASHRAE Standard 70 baseline

data and typical field installations for ceiling diffusers. This data can provide the HVAC

4

design engineer the information necessary to predict diffuser performance in a wide

variety of field installations. The engineer can use the data from this project to adjust

installations to achieve the best performance possible in a given situation. The results can

lead directly to improved system designs with likely benefits to energy efficiency,

occupant comfort and worker productivity.

This report is organized into four chapters covering the step-by-step process taken to

determine the quantitative differences between Standard 70 data and typical field

installations that take into account different inlet conditions. Chapter 2 covers the design

of a diffuser supply air inlet plenum that substantially exceeds Standard 70 requirements

and gives a method for efficient supply plenum design for any throw room. Chapter 3

covers Standard 70 testing using the optimized plenum, where the baseline data for the

output measures (throw, sound and pressure data) were recorded and analyzed for each

type of ceiling diffuser. In Chapter 4, the baseline data was compared to data from field

condition testing for a number of different diffuser inlet combinations. Predictive

algorithms for each output measure based on diffuser inlet conditions were developed.

Chapter 5 covers a comparison of the baseline Standard 70 output measures against data

from a close coupling setup. Finally, conclusions are made and recommendations given

to better improve further testing and real world installations.

5

CHAPTER 2

DIFFUSER SUPPLY PLENUM OPTIMIZATION

An outlet diffuser airflow supply plenum design was experimentally optimized to

minimize spatial variation in airflow velocity at the inlet of the test diffuser. Standard 70

specifies a maximum variation of 10 percent. Variation achieved was a standard

deviation of 2.6 percent, which gives a maximum variation of 5.2 percent for over ninety-

seven percent of the inlet area. To achieve minimal variation in airflow velocity, four test

parameters were varied. There were three for the flow equalization device and one for

plenum air supply inlet configuration. To minimize airflow velocity variation across a

wide spectrum of user conditions, testing was completed across two user selectable

conditions, inlet diameter and volume flow rate. Main effect and interaction results were

analyzed using analysis of variation. This design gives specific build guidance towards

achieving the best plenum performance.

2.1 Background

ANSI/ASHRAE Standard 70-2006 is the accepted method of testing the performance

of air outlets. [9] The standard defines laboratory methods of testing air outlets used to

terminate ducted and non-ducted systems for distribution and return of building air. For

air outlet testing, the standard describes two methods of testing, ducted and plenum. The

ducted method can be considered an ideal installation with a long, vertical rigid duct

leading to the diffuser. The plenum method is a laboratory simulation of an ideal

installation, with the diffuser inlet connected to a large plenum with ideally stagnant air.

The plenum method is often the method of choice because it allows for quick change-out

6

of diffusers with different inlet shapes and sizes, and a smaller requirement for vertical

space above the ceiling in the test facility.

The standard specifies that construction of the plenum should provide uniform and

unidirectional air velocities such that the velocity profile at the entrance plane of the test

device must be within 10 percent at any location. The standard also notes that since

practical considerations will limit the shape and volume of the plenum, equalization

devices may be required to accomplish the flow uniformity but does not specify the

equalization device construction or placement in the plenum. Also, the designer is left to

decide the volume, shape and supply air inlet configuration of the plenum. Thus, the

design of the plenum and accompanying flow equalization device has become an art

based on knowledge of the physics of airflow and experience in test device design.

2.2 Objective

The objective of this experiment was to design and build a plenum that could be used

to characterize diffuser performance, as well as, develop guidance for other designs with

similar flow objectives. The goal for that objective was that the performance of the

plenum should not only meet Standard 70 specifications but come as close as possible to

ideal. The main quality measure was the level of uniformity of airflow at the entrance of

the test diffusers. A secondary objective was to develop and perfect the test methods for

measuring diffuser throw and sound generation for the laboratory and the design of

experimental methods used to test for parameter and noise variation effects. These testing

and experimental design methods were to be used for all the following tests performed in

this project. [10]

7

2.3 Methodology

The one output measure was the airflow velocity across the plane of the diffuser inlet.

The quality characteristic derived from this measure was the variation in airflow velocity

across the plane. To meet the experiment objectives economically, a Taguchi designed

fractional factorial experiment was developed using the four test parameters most likely

to be used for similar use plenums and that would be most likely to affect the output.

Other parameters were held constant during the experiment. To cover the standard range

of test airflow rates and diffuser inlet diameters, testing was performed near the two

extremes of airflows reported in published performance data for two different inlet

diameters.

2.3.1 Experimental Design

The ideal energy transformation is for air to flow into the plenum and exit at the inlet

to the diffuser with a uniform vertical velocity and zero horizontal velocity. Thus, any

variations from ideal would be variations in either direction or airspeed across the plane

of the diffuser inlet. The variations of interest would be large enough and across enough

of the inlet plane to cause some variation in diffuser performance from ideal conditions.

Considering a circular diffuser inlet, a variation in flow velocity that would likely result

in a significant variation in diffuser output velocity would cover an angle of at least 90

degrees. This assumes that less extensive variations would tend to diffuse into the flow

and not be detectable in the output flow. Therefore, an angular frequency of two cycles

for a rotation around the inlet or 1/π (one cycle per 180 degrees) was used. That required

four samples around the inlet plane. These measurements were taken equally spaced,

approximately 1.5 inches inside of the circumference. Across the radius, a variation in

8

flow velocity that would likely result in a significant variation in diffuser output velocity,

would be a difference in velocity between the center and any velocity near the edge.

Therefore, a maximum spatial frequency of 1 (one cycle per diameter) was used requiring

a long any diameter line that traced the circumference of the inlet diameter and had a

center point along with two points, each near opposite ends of the circumference. Thus,

in total, five measurements were made in the inlet plane for each test run. From these

measurements a standard deviation for each run condition was determined.

In product optimization, noise conditions are conditions that are not controlled by the

designer but are set by the product user or result from external factors not controlled by

either the designer or user. The noise conditions considered that could affect the output

were the diffuser inlet size and the volume airflow rate. Tests were conducted with the

volume airflow rate at a low and high level that corresponds to the typical low and high

levels reported in the diffuser performance specifications. Thus, for an 8-inch diameter

inlet, low flow was 200 cfm, high was 800 cfm, and for a 12-inch diameter inlet, low was

400 cfm, high was 1200 cfm.

Test conditions are parameters that the designer can adjust to obtain optimum

performance. For the plenum, three parameters determined the design of the flow

equalization device in the plenum between the top inlet to the plenum and the inlet to the

diffuser. The flow equalization device tested was a circular perforated plate. The flow

equalization device parameters were the distance from the plenum inlet, the size of the

flow equalization disk, and the percentage of open area (due to perforation dimension and

quantity) of the disk. A forth parameter was the ratio of flow from the top to the flow

from two inlets on the sides of the plenum. A list of the test parameters and noise

9

conditions is shown in Table 1. A picture of the plenum exterior showing the inlet

configuration is shown in Figure 1. The plenum has dimensions of 72 inches long by 41

inches high by 45 inches wide. The top inlet duct has a 12-inch diameter, the two side

inlet ducts are 10-inches in diameter. An 18-inch outlet transition piece is placed between

all ducts and the plenum wall. The different types of flow equalization disks and their

sizes are shown in Figure 2.

Obviously, there are many more parameters in the plenum design that could affect the

inlet velocity variation. Among them are plenum size (three dimensions), diameter of

plenum flow inlet ducts, number of plenum flow inlet ducts, design of the inlet cone

attached to the diffuser inlet, and length of straight duct between the diffuser inlet and the

inlet cone. Based on logical argument, a designer can generally assume that increases in

any of those parameters would not result in significant degradation of the results

presented here, and could result in some improvement. In this experiment a small plenum

with relatively small diameter plenum inlet ducts, one to three plenum inlet ducts, and

minimum straight duct between the diffuser inlet and the inlet cone were used, thus

providing an economical value for these parameters while still designing to exceeding

Standard 70 requirements.

10

Table 1. Test Parameter and Noise Condition List for Test Array.

Parameter Low State Mid State High State

1. Inlet air configuration

All from center inlet

Center and side inlets open

Center inlet partially closed

2. Distance of flow equalization to center inlet

4 inches 8 inches 12 inches

3. Size of flow equalization disk

8 inches diameter

12 inches diameter

18 inches diameter

4. Open area of flow equalization disk 50 percent 40 percent 13 percent

Noise Condition Low State High State

1. Diffuser inlet diameter 8 inches 12 inches

2. Volume flow rate Low High

Figure 1. Picture of the test plenum.

11

Figure 2. Pictures of the different types (left) and the different sizes (right) of the flow equalization disks.

The diffuser inlet condition used for the experiment was a round inlet formed by the back

of a standard perforated diffuser with a short inlet cone attached to the diffuser as shown

in Figure 3. The inlet cone is expected to create a smooth transition of the airflow from

zero velocity to the velocity of the diffuser inlet. The core outer diameters were 14 inches

for the 8-inch inlet, 16 inches for the 10-inch inlet, and 18 inches for the 12-inch inlet.

12

Figure 3. Diffuser inlet configuration. 2.3.2 Laboratory Instrumentation

All tests for this project were run under steady-state isothermal conditions. Plenum

volume airflow was set using a variable frequency drive fan motor and measured with an

Ebtron precision airflow/temperature meter in the supply duct with an installed airflow

accuracy of ±3 percent of reading and a repeatability of ±0.25 percent of reading. Since

the output of concern is airspeed variation, the absolute value of the volume airflow (a

noise condition) is not critical so long as it is close to the noise condition. In these tests,

volume airflow within 5 percent of the noise condition level was considered acceptable.

Each of the five measurements, shown in Figure 4, for a test run were an average of

measurements taken every two seconds over at least one minute. Points one through four

were taken approximately 1.5 inches from the edge of the diffuser inlet walls. Airspeed

measurements were taken using a TSI VelociCalc Plus Multi-parameter Ventilation

Meter with a published accuracy of ±3 percent of reading, examples given in Figure 5.

13

Measurements taken for this experiment are used for comparison and not for absolute

values. The replication variation (from one test to another) of the instrument is

anticipated to be less than 1.3 percent.

Figure 4. Diffuser inlet variation testing configuration.

Figure 5. Actual test configuration using TSI Ventilation Meter.

14

2.3.3 Noise Experiment

Prior to performing the optimization experiment, a ‘noise experiment’ was performed,

where the plenum test parameters are set at a nominal level and the noise factors are

varied from high to low to determine the effect of noise levels on the output variation.

The noise factors for this test were diffuser inlet size and volume flow rate. The nominal

design configuration had all four test parameters set at a mid-state. For the noise test, a

full factorial test was performed where all four combinations of diffuser inlet size and

volume flow rate were tested. Airspeed data was taken at the five measurement points

previously described. The five individual airspeed measurements were normalized to the

average of the five measurements. A signal-to-noise ratio (S/N) was calculated for each

test run using the formula,

S/N = -10log10 s2 ,

where s is the sample standard deviation in non-dimensional form of a fraction of mean

airspeed.

Signal-to-noise is a standard parameter used for determining the level of output

variation due to parameters in the test matrix, in this case the two noise conditions. The

higher the signal-to-noise ratio, the smaller the output variation due to noise. The results

are shown in Figure 6. From the results it was determined that variation increases

significantly with size of inlet and slightly with increased airflow. The goal of the

parameter experiment is to determine the parameter levels that achieve the lowest airflow

velocity variation under all the noise conditions. Figure 6 shows that certain levels of the

15

noise conditions result in greater average variation in the noise experiment and would be

expected to cause similar effects in the parameter design experiment. As a result, to

create a robust design where output is less sensitive to noise effects, parameter

optimization testing should include those levels of the noise conditions. Therefore, the

noise condition for the optimization testing was high and low airflow with all test runs at

the largest diffuser inlet diameter.

Figure 6. Signal to noise ratio results of the noise experiment. 2.3.4 Optimization Experiment

A Taguchi Orthogonal Array was used to test the system for all the parameter

variations and the noise condition of the two airflow rates. The L9 array, shown in Table

2, was chosen for this experiment. The array required nine tests and has the ability to

16

evaluate the main effects of four parameters at three levels. The array is balanced by

choosing parameter levels such that any condition of any parameter is tested with an

equal number of high and low conditions of the other parameters. This testing method

reduces interaction effects in the average output and exposes all parameters to the

different levels of the other parameters.

Table 2. Test Array with One 4-Level Parameter, Four 3-Level Parameters And One Noise Parameter at 2 Levels.

Parameters Noise Condition

Level

Test No. Inlet

Configuration

Flow Equalizer

Distance [in] Flow

Restriction

Flow Equalizer

Diameter [in]

Low Flow

400cfm

High Flow

1200cfm 1 1 4 1 8 2 1 8 2 12 3 1 12 3 18 4 2 4 2 18 5 2 8 3 8 6 2 12 1 12 7 3 4 3 12 8 3 8 1 18 9 3 12 2 8

Similar to the noise experiment, the five airspeed measurements were normalized to the

average of the five measurements for each noise condition and a signal-to-noise ratio

(S/N) was calculated for each run from the ten normalized airspeed measurements.

2.4 Results

The information used to analyze the optimization experiment included the main

effects on airspeed variation from each of the four parameters, the interaction between

17

parameters in airspeed variation and the statistical significance of the main effects of each

parameter.

2.4.1 Optimization Experiment

Plots of the main effects for the optimization experiment are shown in Figure 7. The

plots show the signal-to-noise ratio at the three different levels of each parameter.

Figure 7. Main effects plots for the optimization test. These results show that inlet flow condition 3 (flow primarily from the side inlets), flow

equalizer position 3 (18 inches from the top plenum inlet), flow equalizer disk size 3 (18

inch diameter disk), and flow equalizer open area 3 (most restrictive) produced the

highest signal-to-noise ratio, which should equate to the configuration with the lowest

flow velocity variation which is the signal in this case. In other words, this configuration

18

is predicted to have the least variation in output across the inlet area and at the two test

airflow rates.

The significance of the main effects were calculated using an analysis of variatiance.

The results are shown in Table 3 (Note that only two to three digits are significant). The

adjusted sum of the squares (Adj SS) shows how much of the total variation was due to

the corresponding factor. For example, inlet configuration had an adjusted sum of the

squares of 0.000740, the total sequential sum of the squares (Seq SS) was 0.00282, thus

the fraction of the variation due to inlet configuration was 0.00074/0.00282 or roughly

0.26 or 26 percent. The F-statistic shows the ratio of variation due to that factor and the

variation due to noise when taking into account degrees of freedom. The higher the more

significant. The P-value shows the significance of the factor variation in terms of the

probability that this variation could be due to random sampling. Normally, a confidence

level of 95 percent, or P-value of less than 0.05 or 5 percent is needed to consider results

significant. For inlet configuration, the P-value shows a 0.7 percent chance that the

variation was due to random sampling (a 99 percent confidence level). The summary of

this analysis shows that the variation measured for inlet configuration, flow equalization

distance and flow equalization diameter were significant and the variation measured for

flow restriction was marginally significant. This gives valuable information on the levels

of confidence that should be given to the results.

19

Table 3. Analysis of variation of standard deviations for the optimization test.

Source DF Adj SS Adj MS F P Inlet Configuration 2 0.0007396 0.0003698 9.24 0.007 Flow Equalizer Dist_[in] 2 0.0006198 0.0003099 7.74 0.011 Flow Restriction 2 0.0003064 0.0001532 3.83 0.063 Flow Equalizer Diameter_[in] 2 0.0008012 0.0004006 10.01 0.005 Error 9 0.0003602 0.0000400 Total 17 0.00028237

Before going with these levels, the interaction plots were checked to determine if

there were any strong interactions that would indicate any inaccuracy in a choice based

solely on the main effects. The interaction plots are shown in Figure 8. These plots show

a strong interaction between some parameters. Generally speaking, in an interaction plot,

parallel lines show no interaction, non-parallel but matched increasing or decreasing lines

show moderate interaction, and non-parallel and non-matched increasing or decreasing

lines show strong interaction. For example, for interaction between inlet configuration

and flow equalization distance, the top left plot, we see a non-parallel and non-similar

increasing or decreasing lines showing a strong interaction between inlet configuration

and flow equalizer distance. At flow equalization distance of 4, inlet configuration 1 has

the highest standard deviation, but at an flow equalization distance of 12, inlet

configuration 1 has the lowest. This and many of the other interactions could be

anticipated on physical grounds since the flow equalization device was primarily inline

with the main plenum inlet and thus would be expected to have its greatest effect on inlet

configuration 1 and only a small effect on inlet configuration 3.

20

Figure 8. Interaction plots for the optimization test.

2.5 Analysis

For the optimum configuration predicted as inlet configuration 3, flow equalization

distance 12, flow restriction level 3, and flow equalization disk size 18, the model

predicted a standard deviation of 0.31 percent. Note however that the interaction plots

showed that although the lowest average standard deviation for the levels of inlet

configuration was level 3, the lowest single value of standard deviation occurred with

inlet configuration 1, when the other parameters were at the level predicted to be the

optimum. Because of the high interaction levels, verification testing was done for inlet

configuration at all three levels, while the other three parameters were held at the

predicted optimum levels. The predicted and measured results are shown in Table 4.

These verification tests show that although all actual measured standard deviations were

21

well within Standard 70 guidelines, they were higher than predicted by the model. Also,

instead of plenum inlet configuration 3, configuration 2 had the lowest standard

deviation. These inaccuracies in the model can be attributed to the interaction between

parameters. The experimentally verified optimum configuration, test number 2 in Table

4, had a standard deviation of 0.8158 percent. This means that any value must be off by

slightly over six standard deviations to reach the limit of ±5 (6*0.8158 = 4.8948) percent

variation. This happens less than one time per million.

Table 4. Verification testing results.

Test Parameter Configuration Total s Test

Number Inlet

configuration

Flow equalization distance [in]

Flow restriction

Flow equalization diameter [in]

Standard Deviation

[%] 1 1 12 3 18 1.3763 2 2 12 3 18 0.8158 3

(predicted optimum)

3 12 3 18 2.6261

Through plenum optimization testing, it was determined that by having the inlet cone

directly on the diffuser inlet gave the least among of variance in flow at the diffuser inlet.

2.5.1 Performance Variation Effects Due to Inlet Duct Length

It was later determined that for a perforated diffuser with an 8-inch inlet that had 0-

inches of added height from cone to diffuser inlet lead to significantly decreased distance

in throw from the diffuser when compared to testing using the vertical duct method of

Standard 70. Further investigation showed that a section of 7-10 inches of straight duct

added to the 8-inch diffuser inlet, as shown in Figure 9, resulted in diffuser output

identical to using the ducted method. Table 5 shows that, experimentally, the addition of

22

the straight duct onto the 12-inch inlet resulted in slightly higher standard deviation of

flow velocity across the inlet of the diffuser.

Figure 9. Duct height added from diffuser inlet to inlet cone. Table 5. Effect of added height variations for 12-inch diffuser inlet.

Added height between Cone

and Diffuser Inlet Location 15 [in] 10 [in] 4 [in] 0 [in]

E [1] 1.00616 0.992602 0.99046 1.008282 S [2] 0.982781 0.978368 0.976613 1.00112 W [3] 0.977307 0.979422 0.9947 1.006629 N [4] 0.984915 1.002091 0.997462 0.997264

MID [5] 1.048838 1.047517 1.040764 0.986704 STDEV 0.02942 0.028321 0.024157 0.008632 STDEV

CIR 0.012659 0.011344 0.009257 0.005069

23

2.5.2 Performance Variation Effects Due to Inlet Cone Design

One improvement to flow equalization might be to have a cone with a smooth rate of

increase in cross-sectional area such as an exponential horn. This report used a linear

cone, with a constant rate of cross-section area increase. That cone was very inexpensive

to obtain, yet may have caused flow variation at the diffuser inlet. Figure 3 shows the

linear cone as it would be installed directly on the diffuser inlet.

24

CHAPTER 3

BASELINE DIFFUSER CHARACTERIZATION

Ideal diffuser performance data was collected experimentally for six typical ceiling

diffuser types using primarily an airflow supply plenum in accordance with ASHRAE

Standard 70. Testing covered two diffuser parameters, diffuser type and inlet diameter,

and one system parameter, diffuser inlet neck velocity. From the experimental data,

diffuser throw, sound power, and pressure differential across the diffuser was determined

for each test condition. These results were used as a baseline for comparison against

results from typical field installation conditions.

3.1 Background

ANSI/ASHRAE Standard 70-2006 is the accepted method of testing the performance

of air outlets. [9] The standard defines laboratory methods of testing air outlets used to

terminate ducted and non-ducted systems for distribution and return of building air. For

air outlet testing, the standard describes two methods of testing, ducted and plenum. The

plenum method was the primary method used in these experiments because it allows for

quick change-out of diffusers with different inlet sizes and shapes, and in some cases a

smaller requirement for vertical space above the ceiling in the test facility. Details of the

plenum design and qualification were given previously in Chapter 2. For one case (12-

inch inlet, round diffuser) the ducted method was used because the diffuser size was too

large for the plenum outlet. As a crosscheck of the two methods, a few diffusers were

tested using both plenum and ducted methods.

25

3.2 Objective

The objective was to obtain diffuser throw, sound power, and backpressure

performance data when operating at what is considered ideal installation conditions. The

data was collected at three different flow rates on eighteen different diffusers (six

different types of ceiling diffusers, each at three different inlet neck sizes). This data was

the basis of comparison against real world field installation conditions that are examined

in Chapters 4 and 5.

3.3 Methodology

Output measures and derived output measures included:

1. Room 1/3 octave band sound power level and resulting room NC.

2. Pressure difference between the inside of plenum and inside test room.

3. Diffuser throw distance from center of diffuser.

Sound power level in the test room was measured at 1/3 octave bands from 25 to

10,000 Hz. From a subset of those levels, the octave band sound power levels from 125

to 4,000 Hz is calculated, which are then used to determine the room Noise Criteria (NC)

level based on ANSI/ASA 12.2 as recommended by ASHRAE. [9] The total pressure in

the plenum is measured according to ANSI/ASHRAE Standard 70-2006 as the difference

between the pressure in an air line with four pressure taps spaced around the perimeter of

the inside of the plenum and the pressure in the throw room. The static pressure in both

the plenum and the throw room are assumed to be the same as the total pressure in each

space. In the few cases where a vertical duct was used for the diffuser inlet instead of the

26

plenum, the static pressure was measured at three duct diameters from the diffuser inlet.

Total pressure in the duct was determined by summing the static pressure and the velocity

pressure, which was based on inlet size. Draft air velocity is measured with a horizontal

scan from the diffuser using a vertical array of draft sensors. The measurement thus

covered a vertical plane. That plane was normally perpendicular to the edge of the

diffuser, but in a few cases, the maximum throw direction was found to be non-

perpendicular to the diffuser edge. The maximum throw direction was identified from

results of a scan in the vertical plane parallel to the diffuser edge and six feet

downstream. From diffuser throw scans in the maximum throw direction, the maximum

velocity at each distance from the diffuser is the maximum velocity measured in the

vertical plane from which the 150, 100, and 50 fpm throw distances are determined.

3.3.1 Laboratory Instrumentation

A National Instruments LabView virtual instrument was used to monitor and record

airflow, supply air temperature, room temperature, supply static pressure, draft meter

array position and draft meter readings. Sound level measurements were made with all

systems off except for the air supply, boom microphone and the monitoring and recording

computer. Sound measurements were taken with the air supply on and off to record the

diffuser generated sound due to airflow and a background or ambient noise level without

airflow.

3.3.2 Experimental Setup

As described in the previous chapter, the plenum was designed so that the flow

entering the diffuser inlet would be symmetrical with a maximum of 10 percent variation

at any point measured around the inlet as per Standard 70. The diffuser was installed

27

flush with the suspended ceiling, and ten feet from the nearest wall. There were no

obstructions breaking the plane of the ceiling.

To crosscheck the plenum method used for Standard 70 testing, for several cases, a

replicate test was conducted using the vertical ducted method described in the

ANSI/ASHRAE Standard 70-2006. This method called for a minimum vertical inlet duct

length of six diameters and a pressure ring measurement at three diameters from the

diffuser inlet. In this setup, shown in Figure 10, the vertical duct should be sufficiently

long for the flow to stabilize with a uniform velocity cross-section before entering the

diffuser. The results showed that the results from the plenum testing were nearly identical

to results from the vertical duct method.

All tests for this project were run under steady-state conditions. Volume airflow

corresponding to the required inlet velocity was set and allowed to stabilize. Testing was

performed under isothermal conditions.

Figure 10. Standard 70 vertical ducted method.

The ideal energy transformation is for air to flow into the diffuser inlet and exit at the

diffuser discharge with no pressure loss other than velocity pressure, no sound generation

28

and the intended discharge velocity profile. Actual ideal installation measurements show

variations from ideal output due to flow restrictions in the diffuser, turbulence induced in

the flow due to diffuser structure and anomalies in discharge throw due to the geometry

of the diffuser design.

Sound measurements were made using a rotating boom microphone placed in a

location that was determined to be in the reverberant field, meaning the reverberant field

was dominant. Physically, it was as far from the diffuser as possible, near a corner while

maintaining a distance of five feet from either wall as shown in Figure 11.

Figure 11. Boom microphone configuration. The sound signal was captured by the analysis computer and converted to 1/3 octave

band sound levels. In post-processing, the 1/3 octave band levels were transformed to 1/1

octave band levels and corrected for a room background sound level. Then, for each

octave band a room reverberation correction and a standard room correction of minus 10

dB, to convert from sound power to sound pressure, were applied to the noise criteria

curves to obtain the noise criteria level (NC) for the sound generated by the discharge. [3]

The room reverberation correction was an average of corrections derived from both the

time it takes for the sound pressure level in a room to decrease 60 dB and the measured

29

sound level in the room with a calibrated source. It was recommended by ASHRAE, in

Standard 70, that the 1/1 octave band levels of most significance were: 125, 250, 500,

1000, 2000 and 4000 Hz center frequencies. [9] The 1/3 octave band frequencies that

make up the 125 Hz 1/1 octave band frequency are 100, 125 and 160 Hz. A simple

equation was used to take the sound pressure level averages from each 1/3 octave band

frequencies and average them into the 1/1 octave band center frequency sound pressure

level, Lp, of 125 Hz, measured in decibels. Table 6 shows the other 1/1 center

frequencies and the corresponding 1/3 octave band frequencies suggested in Standard 70.

Lp(1/1,125 Hz) =10log10(10Lp(1/3,100 Hz)/10 + 10Lp(1/3,125 Hz)/10 + 10Lp(1/3,160 Hz)/10)

Table 6. 1/1 and 1/3 octave band center frequencies.

1/1 Octave Bands 1/3 Octave Bands 125 Hz 100,125,160 Hz 250 Hz 200,250,315 Hz 500 Hz 400,500,630 Hz 1000 Hz 800,1000,1250 Hz 2000 Hz 1600,2000,2500 Hz 4000 Hz 3150,4000,5000 Hz

The noise criterion curves produced by Beranek specify maximum sound levels permitted

in each octave band for a specific NC curve. [11] Algorithms based off the NC curve

levels for each center frequency determined the overall NC level for each test.

A vertical array of draft meters, as shown in Figure 12, was used to record diffuser

throw velocities at varying distances from the diffuser. The height of each sensor is listed

in Table 7 where TV17 is the top sensor. The draft sensor vertical array was scanned in

30

the direction of maximum flow velocity on the diffuser side that had the longest run

distance from the center of the diffuser. This gave the greatest amount of useful data that

could be used for comparison with the future field installations.

Figure 12. Array of draft meters.

31

Table 7. Vertical draft meter sensor locations.

Sensor # Height from

Floor [in] Height from Ceiling [in]

TV17 107.25 0.75 TV16 104.75 3.25 TV15 102.5 5.5 TV14 100 8 TV13 97.75 10.25 TV12 95.5 12.5 TV11 93 15 TV10 90 18 TV9 85 23 TV8 78.75 29.25 TV7 73.75 34.25 TV6 67.75 40.25 TV5 56 52 TV4 44 64 TV3 31 77 TV2 25 83 TV1 16 92

Inlet duct velocity was the noise condition for the Standard 70 testing. Inlet duct

airflow velocities used were 1200, 800, and 500 fpm. These velocities cover a large span

of typical diffuser flow conditions, including what is typically seen in actual installations.

For the test array, this parameter was considered a noise condition because it is not

controlled by the installation. Under modern systems, where a VAV unit of some type is

used, a typical diffuser will be required to perform under large variations in airflow.

Two design parameters were used in the performance characterization experiment.

They were diffuser type and diffuser inlet diameter. The design parameters and noise

condition with corresponding levels are shown in Table 8.

32

Table 8. Test parameters and noise conditions with corresponding states.

Parameter State 1 State 2 State 3 State 4 State 5 State 6

1.Diffuser Type Square Plaque Perforated

Round Neck

Modular Core

Perforated Round Louvered

2.Inlet Diameter 8 inches 10 inches 12 inches

Noise Condition

Low State

Medium State High State

1.Diffuser inlet

velocity 500 fpm 800 fpm 1200 fpm

A full factorial array was used to set up the experiment. The experimental outputs were

used to extract the main effects of the test parameters and the variation of those main

effects due to the noise conditions. The test array has 18 runs, one parameter at six levels

and one parameter at three levels. There were three sets of output measures, one for each

noise condition. The array is shown in Table 9. Note that for diffusers with a square inlet

(modular core and louvered), a round to square adaptor was used.

From the diffuser velocity profile data, the data was normalized so that diffusers of

the same type with different inlet sizes, at differing velocities, could be compared side-

by-side. To do this, the zone plot method of the diffuser velocity data was used. This

method is described in Appendix C of ANSI/ASHRAE Standard 70-2006 and is

described by the zones of expansion from an air outlet. Using this method, the non-

dimensional diffuser discharge velocity is plotted corresponding to the non-dimensional

measurement distance. The non-dimensional discharge velocity is the ratio of discharge

velocity to inlet duct air velocity, Vx/Vk. The non-dimensional measurement distance is

the measurement distance divided by the square root of the diffuser neck area, X/(Ak)1/2.

33

The results are plotted on a log-log plot and a linear regression curve is drawn to pass

through the data points. For any diffusers characterized by this regression curve, all data

points should fall within ±20 percent of the line as shown in Figure 13. [9]

Table 9. Test array for Standard 70 testing with output measures.

Throw Data at corresponding velocity

[ft] Run #

Diffuser Type

Inlet Size [in] 150fpm 100fpm 50fpm

Sound Data [dB]

Total Pressure

Data [in- H2O]

1 Square 8 2 Square 10 3 Square 12 4 Plaque 8 5 Plaque 10 6 Plaque 12 7 Perf Rnd 8 8 Perf Rnd 10 9 Perf Rnd 12 10 Mod Core 8 11 Mod Core 10 12 Mod Core 12 13 Round 8 14 Round 10 15 Round 12 16 Louvered 8 17 Louvered 10 18 Louvered 12

The log-log plot in Figure 13 shows a zone linear regression line and the

corresponding ±20 percent error lines. This line has been split into three zones labeled 2,

3 and 4. These are standard zones that have a slope based on typical discharge flow

velocity phenomena. Zone 1 (not shown in Figure 13) is known as the short zone and was

rarely seen during analysis of the experimental data. Zone 2 is called the transition zone

34

and can extend eight to ten diameters from the air outlet. Zone 3 is the zone of fully

established turbulent flow and is usually the region that reaches the occupied zone

making its importance the greatest. Zone 4 is the terminal zone meaning the residual

velocity decays into large-scale turbulence at a rapid rate. [12] The regression equation

for zone 1 is y = a*x(0) + b, zone 2 is y = a*x(-1/2) + b, zone 3 is y = a*x(-1) + b and zone 4

is y = a*x(-2) + b. After taking the log of both sides of each equation, notice that the

exponent of x is actually the slope of the regression line in each zone, the value a is the y

value at x = 1 and b is ideally zero. The value of x refers to the y-axis value, Vx/Vk, and

the value of y refers to the x-axis value, X/(Ak)1/2.

Figure 13. Standard 70 Zone Plot.

0.01

0.1

1

1 10 100

Vx/V

k

X/Sqrt(Ak)

Linear Zone Regression Plus 20% Minus 20% Zone 2

Zone 3

Zone 4

35

3.4 Results

Data from the square diffusers are presented as an example of the testing and analysis

performed on all the diffusers. For all test runs, draft meter measurements were used to

plot the diffuser discharge velocity versus distance from the diffuser at various heights

below the ceiling. Figure 14 and Figure 15 show the results for a 12-inch square diffuser

and an 8-inch square diffuser both at 1200 fpm inlet flow velocity. The plot line labeled

TV17-v is the top draft sensor closest to the ceiling at 0.75 inches from the ceiling, while

TV16-v is 3.25 inches from the ceiling, as stated in Table 7. The sensors show that the

flow leaving the diffuser outlet is primarily confined to the space near the ceiling. The

flow spreads out and mixes with air in the room resulting in the velocity slowly dropping

off over a distance of about seven to nine feet. This is a typical ceiling diffuser velocity

profile. From this data comes the throw data distance from the diffuser center for the

velocity points of 150, 100, and 50 fpm.

Figure 14. Velocity profile of 12 inch square diffuser at 1200 fpm.

0 200 400 600 800

1000 1200 1400 1600 1800 2000

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Velo

city

[ft/m

in]

Distance From Diffuser Center [ft]

TV17-v TV16-v TV15-v TV14-v TV13-v TV12-v TV11-v TV10-v TV9-v TV8-v

36

Figure 15. Velocity profile of 8 inch square diffuser at 1200 fpm.

Figure 16 shows the calculated room noise criterion (NC) level for a 12-inch square

diffuser at 1200 fpm for forty samples. In all cases, there were a minimum twenty

samples taken over a two minute time period for each diffuser at each of the three noise

conditions (flow rates). Samples were excluded from the average if there was evidence of

a temporary background noise above the background noise level. The plenum total

pressure was recorded every four seconds during the draft meter flow testing as shown in

Figure 17 and averaged for each test run. Table 10 through Table 15 show the final

Standard 70 results for every diffuser type and inlet size tested for diffuser throw distance

data at 150/100/50 feet per minute, total pressure data inside the plenum in inches of

water, and the room noise criterion (NC) level in decibels.

0

200

400

600

800

1000

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Velo

city

[ft/m

in]



37

Figure 16. Noise criterion level for 12 inch square diffuser at 1200 fpm.

Figure 17. Plenum total pressure for 12 inch square diffuser at 1200 fpm.

37.0

38.0

39.0

40.0

41.0

0 5 10 15 20 25 30 35 40

Dec

ibel

Lev

el (d

B)

Sample Number

NC 12in Square

0.2315

0.2320

0.2325

0.2330

0.2335

0.2340

0.2345

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Tota

l Pre

ssur

e [in

-H2O

]


Plenum Total Pressure (P4)

38

Table 10. Final results for square diffuser at corresponding velocities.

Neck Size [in] Neck Velocity

Manf. AK [ft2] [fpm] 500 800 1200

Flow [cfm] 175 280 420 Total Pressure [in-wt] 0.0322 0.0652 0.1301

Throw [ft] 3 4 9 5 6 10 6 8 13 8

0.3491 NC [dB] - <21 33


Throw [ft] 0 0 0 0 0 0 0 0 0 10

0.5454 NC [dB] <21 24.5 36


Throw [ft] 0 0 0 7 10 15+ 10 13 15+

A

12 0.7854

NC [dB] <21 28 38.5 Flow [cfm] 273 436 654

Total Pressure [in-wt] - 0.08 0.169 C 10 0.5454 NC [dB] <21 22 36

Table 11. Final results for plaque diffuser at corresponding velocities.


Manf. AK [ft2] [fpm] 500 800 1200


Throw [ft] 3 4 8 4 6 12 6 8 15 8

0.3491 NC [dB] - <21 28

Flow [cfm] 273 436 654 Total Pressure [in-wt] 0.048 0.102 0.213 10

0.5454 NC [dB] - <21 32.5 Flow [cfm] 393 628 942

Total Pressure [in-wt] 0.05829 0.137448 0.325853 Throw [ft] 5 8 12 8 11 14 12 15 15+

C

12 0.7854

NC [dB] <21 23.5 36 Flow [cfm] 273 436 654

Total Pressure [in-wt] - 0.122 0.269 A 10 0.5454 NC [dB] <21 27 39.5

39

Table 12. Final results for perforated round diffuser at corresponding velocities.


Manf. AK [ft2] [fpm] 500 800 1200


Throw [ft] 1 1 1 1 1 2 1 1 2 8

0.3491 NC [dB] <21 23.5 37


Throw [ft] - - - - - - 6 9 14 10

0.5454 NC [dB] <21 32 46.5


Throw [ft] 4 5 12 6 8 15 9 13 15+

B

12 0.7854

NC [dB] <21 36 52 Flow [cfm] 273 436 654

Total Pressure [in-wt] 0.04677 0.102477 0.232094 Throw [ft] 1 1 3 2 3 4 3 4 6 C 10

0.5454 NC [dB] <21 32.5 46.5

40

Table 13. Final results for perf square/modular core diffuser at corresponding velocities.


Manf. AK [ft2] [fpm] 500 800 1200


Throw [ft] 3 4 7 4 6 10 5 7 12 8x8

0.3491 NC [dB] <21 30 42

Flow [cfm] 273 436 654 Total Pressure [in-wt] 0.04428 0.097157 0.217667 10x10

0.5454 NC [dB] <21 30.5 44.5 Flow [cfm] 393 628 942

Total Pressure [in-wt] 0.04617 0.110369 0.268157 Throw [ft] 4 5 10 6 9 15 10 14 15+

D

12x12 0.7854

NC [dB] <21 33 49 Flow [cfm] 175 279 419

Total Pressure [in-wt] 0.03309 0.066259 0.132328 Throw [ft] 1 1 4 2 3 6 3 5 8

8x8 0.3491

NC [dB] - <21 33 Flow [cfm] 273 436 654

Total Pressure [in-wt] - - 0.187947 Throw [ft] - - - - - - 6 9 15

B

10x10 0.5454

NC [dB] - - 42.5

41

Table 14. Final results for round diffuser at corresponding velocities.


Manf. AK [ft2] [fpm] 500 800 1200


Throw [ft] 3 4 7 4 6 11 6 9 14 8

0.3491 NC [dB] - <21 25.5

Flow [cfm] 273 436 654 Total Pressure [in-wt] - - 0.154607

Throw [ft] - - - - - - 8 12 15+ 10

0.5454 NC [dB] - <21 29


Throw [ft] 3 5 7 5 8 11 9 12 15+

C

12 0.7854

NC [dB] <21 25 38.5 Flow [cfm] 273 436 654

Total Pressure [in-wt] 0.03871 0.082677 0.185251 Throw [ft] 3 5 7 4 6 10 6 10 15+ B 10

0.5454 NC [dB] <21 24.5 36.5

42

Table 15. Final results for louvered diffuser at corresponding velocities.


Manf. AK [ft2] [fpm] 500 800 1200


Throw [ft] 5 7 12 7 10 15+ 10 14 15+ 9x9

0.3491 NC [dB] - <21 29.5


Throw [ft] 5 8 14 8 10 15+ 10 15 15+ 12x12 0.5454

NC [dB] - <21 34 Flow [cfm] 393 628 942

Total Pressure [in-wt] 0.02845 0.062111 0.142311 Throw [ft] 6 8 15 8 12 15+ 11 15 15+

D

15x15 0.7854

NC [dB] - <21 34 Flow [cfm] 169 279 394

Total Pressure [in-wt] 0.054089 - 0.21424 Throw [ft] 5 8 14 - - - 10 15 15+

9x9 0.3491

NC [dB] <21 - 38.5 Flow [cfm] 393 628 942

A

15x15 0.7854 NC [dB] <21 23.5 38

Note that all diffuser manufacturers’ products may not have the same performance and

the above data may or may not coincide with Standard 70 data published by all

manufacturers.

Figure 18 shows a typical example zone plot of throw data for a type of diffuser, in

this case, all the square diffusers tested. The data includes runs with different neck sizes

and manufacturers at several different duct air velocities. What can be seen from the plot

data is that the non-dimensional data for the different diffusers nearly overlap one

another. This shows the value of the zone plot for using data taken at a few neck sizes

and duct air velocities. This data can be used to predict diffuser performance at different

neck sizes and velocities without having to test at every neck size and flow velocity.

43

Figure 18. Standard 70 zone plot for all square diffusers.

3.5 Analysis

This section covers the analysis of each type of diffuser examined for Standard 70

data. Noise criterion data was not attainable at levels below 21.5 dB due to the

background noise level in the throw room. Therefore, at the low duct air velocity

condition for most diffusers, there is no value.

3.5.1 Square Diffuser

As shown in the zone plot in Figure 18, it is apparent that for all manufacturer

diffusers and all inlet sizes, the throw data fits within the ±20 percent margins specified

by Standard 70. Sound data is nearly identical based on inlet size and manufacturer.

0.01

0.1

1

10

1 10 100

Vx/V

k

X/Sqrt(Ak)

Throw/TerminalVel A 8" 1200 [fpm] Throw/TerminalVel A 8" 800 [fpm] Throw/TerminalVel A 8" 500 [fpm] Throw/TerminalVel A 10" 1400 [fpm] Throw/TerminalVel A 10" 900 [fpm] Throw/TerminalVel A 10" 400 [fpm] Throw/TerminalVel A 12" 1200 [fpm] Throw/TerminalVel A 12" 800 [fpm] Throw/TerminalVel A 12" 500 [fpm] Throw/TerminalVel C 10" 900 [fpm] Throw/TerminalVel D 10" 900 [fpm] Linear Zone Regression Plus 20% Minus 20%

44

There are some minor differences in the total pressure for differing diffuser

manufacturers. This lead to the conclusion that the differences in manufacturer design of

the square diffuser resulted in small enough effects in the performance of the diffuser

output that the same performance zone plot could be used to predict diffuser performance

(see Table 10).

3.5.2 Plaque Diffuser

The throw data for the plaque diffusers is much like that of the square diffusers.

Despite the difference in manufacturer designs, the projections of flow from the output of

the diffusers are again closely comparable within the ±20 percent margins. This does not

hold true for the noise data. The difference in diffuser design has resulted in changes in

NC levels greater than 3 dB between manufacturers, leading to differences in total

pressure (see Table 11).

Table 16. Plaque diffuser total pressure and sound comparisons of Standard 70 test results from manufacturer published data and UNLV throw room data.

Manufacturer Data UNLV Data

Sound [dB]

Pressure [in-H2O]

Sound [dB]

Pressure [in-H2O]

800 fpm 26 0.101 27 0.122 A , 10 “ 1200 fpm 37 0.229 39.5 0.269 800 fpm 22 0.115 <21.5 0.103 C , 10 “ 1200 fpm 36 0.259 32.5 0.216

45

Figure 19. Standard 70 zone plot for plaque diffusers. 3.5.3 Perforated Round Diffuser

The maximum throw direction for the round inlet, perforated face diffusers were not

always perpendicular to the diffuser edge. A cross flow scan was used to find the peak

throw direction for data collection. Figure 20 shows the data for two different

manufacturers at various inlet sizes and duct air flow velocities. It is clear that the

differences in the two manufacturer’s designs result in unique zone plots. Much like the

square diffuser, the sound and pressure data for the perforated round diffuser were almost

identical for corresponding inlet sizes and manufacturers (see Table 12).

0.01

0.1

1

10

1 10 100

Vx/

Vk

X/Sqrt(Ak)

Throw/TerminalVel C 12" 1200 [fpm] Throw/TerminalVel C 12" 800 [fpm] Throw/TerminalVel C 12" 500 [fpm] Throw/TerminalVel A 10" 1400 [fpm] Throw/TerminalVel A 10" 900 [fpm] Throw/TerminalVel A 10" 400 [fpm] Throw/TerminalVel C 8" 1200 [fpm] Throw/TerminalVel C 8" 800 [fpm] Throw/TerminalVel C 8" 500 [fpm] Linear Zone Regression Plus 20% Minus 20%

46

Figure 20. Standard 70 zone plot for perforated round diffusers. 3.5.4 Modular Core/Perforated Square Diffuser

The data for manufacturer D used a perforated square diffuser that has a set of

directional veins fixed inside the diffuser inlet, giving the diffuser the name modular

core. The data for manufacturer B used a normal perforated square diffuser with an open

inlet. The perforated square, modular core, diffusers from manufacturer D performed

nearly identical based on inlet size. Regular perforated square diffusers were used from

manufacturer B, each inlet size for the differing perforated square types does not fall

within the ±20 percent margins for the modular core diffuser of manufacturer D. As

noted earlier, performance data developed in this investigation did not always match

0.01

0.1

1

10

1 10 100

Vx/V

k

x/Sqrt(Ak)

Throw/TerminalVelocity B 12" 1200 [fpm] Throw/TerminalVelocity B 12" 800 [fpm] Throw/TerminalVelocity B 12" 500 [fpm] Throw/TerminalVelocity B 8" 1200 [fpm] Throw/TerminalVelocity B 8" 800 [fpm] Throw/TerminalVelocity B 8" 500 [fpm] Throw/TerminalVelocity B 10" 1200 [fpm] Throw/TerminalVelocity B 10" 400 [fpm] Throw/TerminalVelocity C 10" 1200 [fpm] Throw/TerminalVelocity C 10" 800 [fpm] Throw/TerminalVelocity C 10" 500 [fpm] Linear Zone Regression Plus 20% Minus 20%

47

manufacturer published data. In this case, the NC levels are approximately 6 dB above

that of the manufacturer’s published data for the modular core 8x8” inlet size, and the

total pressure was approximately twice the published data (see Table 13).

Table 17. Modular core total pressure and sound comparisons of Standard 70 test results from manufacturer published data and UNLV throw room data.

Manufacturer Data UNLV Data Sound [dB] Pressure Sound [dB] Pressure

500 fpm -- 0.02 -- 0.0533 800 fpm 24 0.05 30 0.108 D , 8x8 “ 1200 fpm 36 0.11 42 0.202

Figure 21. Standard 70 zone plot for perforated round/modular core diffusers.

0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)

Throw/TerminalVelocity B 8x8 1200 [fpm] Throw/TerminalVelocity B 8x8 800 [fpm] Throw/TerminalVelocity B 8x8 500 [fpm] Throw/TerminalVelocity B 10x10 1200 [fpm] Throw/TerminalVelocity B 10x10 400 [fpm] Throw/TerminalVelocity D 8x8 ModCore 1200 [fpm] Throw/TerminalVelocity D 8x8 ModCore 800 [fpm] Throw/TerminalVelocity D 8x8 ModCore 500 [fpm] Throw/TerminalVelocity D 12x12 ModCore 1200 [fpm] Throw/TerminalVelocity D 12x12 ModCore 800 [fpm] Throw/TerminalVelocity D 12x12 ModCore 500 [fpm] Linear Zone Regression Plus 20% Minus 20%

48

3.5.5 Round Diffuser

The round diffusers behave much like the square diffusers. All throw data fits within

the ±20 percent margins of the zone plot regression lines. Note also that the NC levels for

some manufacturers were much less than what was previously published, in some cases,

10 dB or more. Other manufacturers follow very closely to published data (see Table 14).

Figure 22. Standard 70 zone plot for round diffusers.

0.01

0.1

1

10

1 10 100

Vx/

Vk

X/Sqrt(Ak)

Throw/TerminalVel B 10" 1200 [fpm] Throw/TerminalVel B 10" 800 [fpm] Throw/TerminalVel B 10" 500 [fpm] Throw/TerminalVel C 10" 1200 [fpm] Throw/TerminalVel C 10" 400 [fpm] Throw/TerminalVel C 8" 1200 [fpm] Throw/TerminalVel C 8" 800 [fpm] Throw/TerminalVel C 8" 500 [fpm] Linear Zone Regression Plus 20% Minus 20%

49

3.5.6 Louvered Diffuser

Again, just like the square diffuser, the louvered diffuser performs almost identical for

each inlet size and brand. All throw data fits within the ±20 percent margins of the zone

plot linear regression lines (see Table 15).

Figure 23. Standard 70 zone plot for louvered diffusers. 3.5.7 Performance Variation Effects Due to Non-Perpendicular Flow

It was later determined that diffuser throw could be skewed, in a non-perpendicular

direction form the diffuser face despite the equalized flow entering the diffuser inlet.

Cross flow scans, parallel to the diffuser face, showed exactly where the maximum throw

0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)

Throw/TerminalVelocity D 15x15 1200 [fpm] Throw/TerminalVelocity D 15x15 800 [fpm] Throw/TerminalVelocity D 15x15 500 [fpm] Throw/TerminalVelocity A 9x9 800 [fpm] Throw/TerminalVelocity A 9x9 500 [fpm] Throw/TerminalVelocity D 9x9 1200 [fpm] Throw/TerminalVelocity D 9x9 800 [fpm] Throw/TerminalVelocity D 9x9 500 [fpm] Throw/TerminalVelocity D12x12 1200 [fpm] Throw/TerminalVelocity D12x12 800 [fpm] Throw/TerminalVelocity D12x12 500 [fpm] Linear Zone Regression plus20% minus 20%

50

from the diffuser occurred. A similar scan was then performed on the opposite side. If

necessary, cross-flow scans of all four sides were performed to get a total picture of the

diffuser discharge flow. A top view of a typical cross flow scan pattern for the forward

side of a diffuser is shown in Figure 24. Once the direction for the maximum throw was

determined, a linear scan was completed in that direction. Note that based on the

direction of maximum throw, the linear scan could traverse in directions non-

perpendicular to the diffuser side.

Figure 24. Cross flow scan pattern for the forward side of a diffuser.

51

One case was found during verification testing of the 8” perforated round diffuser

using the vertical ducted method from Standard 70 where a cross flow scan was

performed to pinpoint the direction of the peak throw velocity. Figure 25 shows that the

peak throw velocity occurs six inches to the right of the diffuser center at a distance of six

feet. The linear scan of the diffuser throw was adjusted to the proper angle to capture the

maximum throw.

Figure 25. Cross flow scan of 8” perforated round diffuser using vertical ducted method.

0

40

80

120

160

-7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7

Velo

city

[ft/m

in]



52

CHAPTER 4

FIELD CONDITION AIR OUTLET PERFORMANCE

Laboratory full scale testing of diffuser performance, when installed with the typical

variation of field installation practices was conducted and compared to the results of

Standard 70 characterization. The installation variations included type of duct going to

the diffuser, approach direction of the duct, vertical duct height going into the diffuser

intake, and use of a damper in the intake. The airflow total pressure loss, discharge

symmetry and noise generation were all affected by the variations in installation.

Invariably, the standard 70 characterizations had the lowest pressure loss, least discharge

asymmetry and lowest noise generation, although some field installation approached

Standard 70 levels. The increases in these output measures were quantified by

comparison to the output levels measured in a Standard 70 installation. Installers and

engineers, to predict the actual performance of a variety of field installations, can use this

quantitative information.

4.1 Background

Manufacturers of ceiling diffusers publish values of throw, pressure (static and total)

and noise criteria levels (NC) based on testing conducted according to Standard 70 the

accepted method of testing the performance of air outlets. [9] However, under typical

field conditions, diffusers and outlets are seldom installed with the same conditions as

measured under ASHRAE Standard 70. Installation variations that can result in

significant performance variation from ASHRAE Standard 70 installation include the

53

length and type of the duct branch, how duct turns are accomplished and how the duct

approaches the diffuser.

The duct approach to the diffuser is particularly important, since detrimental effects

of improper duct approach cannot be totally corrected by the diffuser itself. Accepted

guidance states that the velocity of the air stream should be as uniform as possible over

the entire outlet connection to the duct and must be perpendicular to the outlet face.

However, few outlet installations achieve this goal. [3,4]

Manufacturers publish sound levels versus flow rate for a given diffuser installed as

called for in ASHRAE Standard 70. However, duct connections encountered in the field

result in significantly different and usually higher sound levels for the same airflow rate.

[1] Also, diffusers with a perforated face have been found to have a higher sound

distortion with poor inlet conditions than diffusers with large open cones.

The magnitude of the sound and performance effects associated with installation

variations will also depend on variations out of the control of the installer such as

variable flow rates. Diffusers are designed to optimally distribute the air across a limited

range of load condition and air volume, but in a typical VAV type installation, air volume

rate can have large variation. Consequently, the throw, room airflow velocity and sound

levels can be significantly different from the specified design point. [1]

4.2 Objective

The objective of this testing was to develop guidelines that will relate manufactures’

air outlet cataloged data that have been obtained using ASHRAE Standard 70 to field

installed application conditions. Data was collected for diffusers obtained under a

54

collection of real world inlet conditions and compared to base performance data obtained

from testing using ASHRAE Standard 70. From that comparison, correction factors for

diffuser throw, pressure and sound data were developed.

4.3 Methodology


1. Sound Power Level and Resulting Room NC.

2. Ratio of total pressure in field installation to total pressure in a Standard 70 test.

3. Diffuser throw distance ratio of throw in field installation to throw in a Standard 70

test for at least two diffuser sides.

Sound power level is the level in decibels of the sound power at third octave bands

from 16 to 10,000 Hz. From a subset of those levels, the room Noise Criteria (NC) levels

were calculated based on ANSI/ASA 12.2 as recommended by ASHRAE. [9] The same

process used during the Standard 70 baseline testing was repeated for the field-testing

sound capture with the rotating boom microphone setup in Figure 11. The throw ratio is

the ratio of the installed throw distance of the specific configuration tested divided by the

throw distance from the standard 70 configuration. Pressure loss is the difference of the

total pressure upstream of the diffuser minus the total pressure in the throw room. The

total pressure in the upstream duct is the summation of the static and dynamic pressures

while the total pressure in the throw room is assumed to be the static pressure in the

throw room. The velocity profile is the two-dimensional profile of the diffuser discharge

velocity in a vertical plane that is parallel to the horizontal duct feeding the diffuser. A

55

profile in the vertical plane perpendicular to the horizontal duct may also be obtained.

The flow factor change is related to the pressure required to achieve a given airflow rate

through a duct or diffuser. In this case, the parameters related to the elbow in the duct, the

damper and diffuser combined system all affect the airflow resistance.


All tests for this project were run under steady-state conditions similar to the Standard

70 tests. Volume airflow corresponding to the required inlet velocity was set and allowed

to stabilize. The throw room measurement instrument configuration was the same as in

the Standard 70 testing. Also, the same LabView virtual instrument was used to monitor

and record airflow, supply air temperature, room temperature, supply static pressure,

draft meter array position and draft meter readings. Sound level measurements were

made with all systems off except for the air supply, boom microphone and the monitoring

and recording computer. Sound measurements were taken with the air supply on and off

to record the diffuser generated sound with airflow and a background noise level without

airflow.



diffuser discharge with no pressure loss, other than velocity pressure, no sound

generation and a discharge velocity pattern as intended by the diffuser design. Thus, any

variations from ideal would be variations in pressure measured in the duct upstream of

the diffuser inlet, increases in sound level measured in the discharge room and

unintended asymmetry in diffuser throw measured in the discharge room. Comparisons

with Standard 70 measurements show variations from ideal due to flow restrictions in the

56

diffuser, turbulence induced in the flow due to diffuser structure and anomalies in

discharge throw due to the geometry of the diffuser design. In this experiment, the

Standard 70 results are used as a standard while the test results are compared to that

standard. The experimental setup is diagrammed in Figure 26.

Figure 26. Experimental setup for field installation condition laboratory testing.

Experimental output measures were obtained using the same or comparable methods

used for standard 70 testing and thus can be compared to Standard 70 measurements.

Static pressure was measured three duct diameters upstream of the elbow that connects to

the vertical duct section leading into the diffuser. Velocity pressure was calculated from

measurement of the total airflow and knowledge of the duct cross-sectional area. Total

pressure was calculated by adding the velocity pressure to the static pressure.

57

Diffuser discharge velocities at incremental distances from the diffuser were

measured using the array of draft meters used in the Standard 70 measurements. The

array of draft meters was scanned in the direction of the greatest discharge. Because the

field installation, in many cases, resulted in non-uniform changes to the discharge

velocity when compared to Standard 70 testing results, care was taken to determine the

location and direction of the highest discharge throw for all sides of the diffuser. This was

accomplished with a cross-flow scan at a set distance from the diffuser side from which

the discharge flow directional characteristics could be determined. Refer to section 3.5.7

in Chapter 3 for details pertaining to cross-flow scans.

For this experiment, inlet duct velocity was the noise condition. Noise conditions, in

product optimization, are conditions that are not controlled by the designer but are set by

the product user or result from external factors not controlled by either the designer or the

user. Diffuser inlet duct airflow velocities used were 1200, 800, and 500 fpm. These

velocities cover a large span of typical diffuser flow conditions including what is

typically seen in actual installations. This parameter was considered a noise parameter,

since it is not controlled by the installation. Under modern systems, where a VAV unit of

some type is used, a typical diffuser installation will be required to perform under large

variations in airflow.

Six parameters were used in the performance characterization experiment. They were

diffuser type, diffuser inlet diameter, vertical duct height directly above the diffuser inlet,

the type of duct, damper configuration in the diffuser inlet, and the angle of approach to

the diffuser side of the horizontal duct run. The parameter and noise configurations with

levels are shown in Table 18.

58

Table 18. Test parameters and noise conditions with different states.

Parameter State 1 State 2 State 3 State 4 State 5 State 6

1. Diffuser Type Square Plaque Perforated

Round Neck

Modular Core

Perforated Round Louvered


3. Vertical Duct

Height

0 duct diameters

1.5 duct diameters

3 duct diameters

4. Type of Duct All rigid

Rigid elbow, flex

duct All flex

5. Damper No damper

Damper parallel to

duct

Damper perpendicular

to duct

6. Approach

Angle 0 degrees 23 degrees 45 degrees

Noise Condition Low State Medium

State High State

1. Diffuser inlet

velocity 500 fpm 800 fpm 1200 fpm

The most economical (least number of test runs) Taguchi fractional factorial mixed

array was used to set up the experiment that could be used to extract the main effects of

the test parameters and the variation of those main effects due to the noise conditions. A

modified L18 array, called an L18 mixed array was used. [10] The test array has 18 runs,

one parameter at six levels and five parameters at three levels. The total degrees of

freedom (DOF) needed for the main effects are 15 and this array has 17 DOF. This was

determined for the main effects by taking the total number of levels in each parameter,

subtracting one from each for the mean, and adding them together. The DOF for the test

array is merely the number of runs minus one for mean to determine variability. The

array is shown in Table 19.

59

Table 19. Test array for field condition variations testing.

Run Order

Run #

Diffuser Type Inlet

Duct Height

Type Duct Damper

Damper Type

Approach Angle

2 1 Square 8 0 Rigid None RndSliding 0 5 2 Square 10 1.5 Mix Perpendicular RndSliding 23

16 3 Square 12 3 Flex Parallel RndSliding 45 12 4 Plaque 8 0 Mix Perpendicular RndSliding 45 8 5 Plaque 10 1.5 Flex Parallel RndSliding 0 1 6 Plaque 12 3 Rigid None RndSliding 23

14 7 Perf Rnd 8 1.5 Rigid Parallel RndSliding 23 6 8 Perf Rnd 10 3 Mix None RndSliding 45

15 9 Perf Rnd 12 0 Flex Perpendicular RndSliding 0 3 10 Mod Core 8 3 Flex Perpendicular SqOpBld 23

17 11 Mod Core 10 0 Rigid Parallel SqOpBld 45 11 12 Mod Core 12 1.5 Mix None SqOpBld 0 13 13 Round 8 1.5 Flex None RndSliding 45 18 14 Round 10 3 Rigid Perpendicular RndSliding 0 9 15 Round 12 0 Mix Parallel RndSliding 23 4 16 Louvered 8 3 Mix Parallel SqOpBld 0 7 17 Louvered 10 0 Flex None SqOpBld 23

10 18 Louvered 12 1.5 Rigid Perpendicular SqOpBld 45

There were three sets of output measures, one for each noise condition. The output

array for the noise condition, 1200 fpm, is shown in Table 20.

Table 20. Output array at high state noise condition for the field installation experiment.

Forward Throw

Asymmetry, Tf

Backward Throw

Asymmetry, Tb

Change in NC [dB]

Total Pressure

Ratio Run # 1200 fpm 1200 fpm 1200 fpm 1200 fpm

1 2 … 17 18

60

All duct leading to the diffuser was round, either rigid or flex. For diffusers with a

square inlet (modular core and louvered), a round to square adaptor was used. Damper

types used where the round sliding and the square opposed blade dampers. Approach

angle is measured from perpendicular to the diffuser side.

4.4 Results

Each diffuser output measure is a comparison of the real-world installation

configurations against the corresponding data from the base Standard 70 tests. The

Standard 70 data used for comparison was the data from the same diffuser type of the

same manufacturer at the same flow rate used in the field installation. The asymmetry

was determined by comparing the throw distance in zone 3 of the zone plots from the

real-world configurations and the Standard 70 tests. Figure 27 shows the velocity profile

of a 12-inch square diffuser at 1200 fpm, which is used to develop the zone plot in Figure

28.

Figure 27. Standard 70 throw data at 1200 fpm for 12” square diffuser.

0 200 400 600 800

1000 1200 1400 1600 1800 2000

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Velo

city

[ft/m

in]



61

To determine the zone throw distance at each recorded data point, a ratio was

determined between the maximum throw velocity at each data point versus the calculated

neck velocity based on the air flow rate and diffuser neck size, or Vx/Vk. This ratio is

used as the y-axis of the zone plot in Figure 28. Then a ratio that accounts for throw

distance from the center of the diffuser versus the neck area was calculated. The square

root of the neck area was taken to linearize the data points into proper zone profiling,

giving X/(Ak)1/2. This ratio was the x-axis of the zone plot in Figure 28 resulting in a log-

log plot deemed a diffuser zone plot. This shows how the discharge air is reacting as it

propagates away from the center of the diffuser. All values are put into units of feet

before becoming non-dimensional ratios.

Referring to Figure 13 for zone plot slope information and the description in Chapter

3, the following analysis described using the zone plot in Figure 28 was performed to

compare Standard 70 throw data to the field installation configuration throw data. This

determined the asymmetry of the field-testing installations.

The zone plots for the field-testing results and the Standard 70 test results were used

to determine the effects on throw due to the field installation. Using the starting x1 and

y1 values for the zone three slope (yellow dot), at a constant y2 value of 0.2 for the throw

velocity over the inlet velocity, which was held constant for each comparison, a value for

x2 (pink dot), throw distance over square root of the inlet area, was calculated. The

equation for slope, ([y2-y1]/[x2-x1]), is used to find the x2 value where the y2 value is

0.2. This method was used to obtain a ratio between this x2 value (Standard 70 test) and

the corresponding x2 value for the field installation test. The ratio is the ratio-of-throw

distance for the two conditions. The y-axis value was held constant at 0.2 because for

62

every test configuration this point was in zone three and so that the comparison of each

configuration could then be related on the same scale for predicting the output. The x2

value was calculated for each standard 70 and field installation in the same manner. From

those values, throw ratios of each field installation noise condition to standard 70,

(x2/x2std70) of the calculated x2 values, were made to show how much throw was

affected by the installation parameters.

Figure 28. Standard 70 zone plot for 12” square diffuser, where (x1,y1) symbolizes start of zone 3 (yellow dot) and (pink dot) gives the x2 value when y2 is 0.2.

0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)

Throw/TerminalVelocity 1200 [fpm] Throw/TerminalVelocity 800 [fpm] Throw/TerminalVelocity 500 [fpm] Linear Zone Regression minus 20% plus20%

0.2

8.46

Zone 2

Zone 3

Zone 4

0.85

2

(x1,y1)

(x2,y2)

63

The change in NC level is a direct difference between the NC of the field-testing

configurations and the NC of the Standard 70 tests. The often, low NC levels at 500 fpm

were not always obtainable during the Standard 70 tests due to the background sound

level of the throw room. Therefore, this data was not considered in the analysis. The total

pressure ratio is determined by comparing the averaged total pressure of the field-testing

configurations against the total pressure of the plenum in the Standard 70 tests.

For example, Figure 29 through Figure 32 show the throw velocities and

corresponding zone plots for both forward and backward throw for run number 3 at 1200

fpm inlet duct velocity. Based on this data and the previously calculated Standard 70

data, the final results for throw asymmetry, sound and pressure ratio were determined.

Figure 29. Field installation run #3 forward throw at 1200 fpm.

0 200 400 600 800

1000 1200 1400 1600 1800 2000

0 1 2 3 4 5 6 7 8 9

Velo

city

[ft/m

in]



64

Figure 30. Field installation run #3 backward throw at 1200 fpm.

Figure 31. Zone plot of field install run #3 forward throw 1200 fpm.

0 200 400 600 800

1000 1200 1400 1600 1800 2000

0 1 2 3 4 5 6 7 8 9

Velo

city

[ft/m

in]



0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)

Throw/TerminalVelocity 1200 [fpm] Linear Zone Regression minus 20% plus20%

65

Figure 32. Zone plot of field install run #3 backward throw 1200 fpm. Table 21 shows the recorded diffuser output throw ratios, noise criteria level changes

and total pressure ratios for field installations compared to Standard 70 data for all test

runs. A statistical analysis was conducted to help better develop diffuser throw, noise

criteria and total pressure ratio prediction models for future field installations. In order to

statistically analyze the output measures, they had to be quantifiable. The statistical tool

used was Minitab. The main effects and interaction plots were determined based on the

parameters listed in Table 18 independently for each output measure listed in Table 20.

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)

Throw/TerminalVelocity 1200 [fpm] Linear Zone Regression minus 20% plus20%

66

Table 21. Diffuser output throw ratios, noise criteria level changes and total pressure ratios for field installations compared to Standard 70 data.

Forward Throw Asymmetry, Tf

Backward Throw Asymmetry, Tb

Change in NC [dB]

Total Pressure Ratio

Run #

500 fpm

800 fpm

1200 fpm

500 fpm

800 fpm

1200 fpm

800 fpm

1200 fpm

500 fpm

800 fpm

1200 fpm

1 1.53 1.53 1.53 0.49

0.49

0.45 8.00

6.50 1.63

1.71 1.78 2 0.91 0.91 0.91 1.0

5 1.05

1.05 11.50

11.50 2.17

2.14 2.25 3 1.42 1.42 1.50 1.5

0 1.50

1.50 8.50

8.50 1.70

1.89 1.85 4 1.04 1.04 1.04 0.5

0 0.47

0.47 8.50

10.5 2.19

2.27 2.40 5 0.69 0.69 0.69 1.3

7 1.37

1.37 15.00

13.5 2.17

2.35 2.54 6 0.96 0.96 1.01 1.0

1 1.01

0.99 6.00

5.50 1.16

1.18 1.19 7 0.97 0.97 0.90 1.2

1 1.06

1.06 9.50

9.50 2.11

2.09 2.18 8 1.03 1.03 1.03 0.9

9 0.99

0.99 2.50

2.00 1.24

1.22 1.23 9 1.12 1.16 1.16 0.5

9 0.81

0.81 11.00

5.50 1.75

1.78 1.73 10 1.28 1.32 1.47 0.9

9 1.07

1.12 2.50

2.50 1.55

1.65 1.46 11 1.30 1.30 1.30 0.9

6 0.96

0.96 4.00

3.00 1.34

1.39 1.35 12 1.13 1.25 1.16 0.9

3 0.88

0.83 2.00

1.00 1.25

1.24 1.18 13 1.04 1.04 1.04 0.8

9 0.89

0.89 7.00

7.00 1.42

1.52 1.61 14 1.26 1.26 1.13 1.3

0 1.30

1.12 10.00

9.50 1.75

1.91 1.87 15 1.56 1.44 1.62 1.3

2 1.29

1.42 9.50

10.00 1.41

1.48 1.50 16 0.98 0.89 0.89 0.9

4 0.84

0.84 5.5 8.00 1.94

2.08 2.18 17 1.32 1.32 1.32 0.1

3 0.14

0.16 12.50

8.50 2.32

2.56 2.59 18 1.37 1.44 1.38 0.5

6 0.70

0.68 11.00

9.50 1.72

1.85 1.84

4.5 Analysis

The main effects show the average value of the measured output at the different levels

of each of the input parameters. The example main effects plots used for data analysis of

forward throw are shown in Figure 33. Forward throw is the output and each plot shows

how the throw varied with each parameter at different levels. The horizontal line gives

the overall averaged data means of the forward throw data and acts as the reference value

to compare the effect of each parameter. Diffuser type and inlet diameter are parameters

that were used during the Standard 70 testing. Duct height, duct type, damper and

approach angle are significant to the field installation testing. In the statistical model, it is

67

assumed that each parameter is an additive parameter to the system mean. However based

on the physics of the system, some of the parameters are known to affect the sensitivity

of the system and become multipliers instead of additive values. In this experiment,

diffuser type and inlet diameter are parameters that add sensitivity to the system. With

different diffuser types, comes differing inlet to outlet ratios, meaning some diffusers

have the same outlet size but changing inlet diameter, while others increase or decrease

outlet size along with increased or decreased inlet diameter.

Figure 33. Main effects plot for field installations forward throw.

The example interaction plots for data analysis of the forward throw are shown in

Figure 34. Each plot shows the average output level for the different combinations of two

parameters. If the two variables have an interaction, which is shown when the lines cross

68

or are not parallel between points, then their effects on the output are not simply additive.

Otherwise, if the lines between each point in Figure 34 are parallel, the output effects of

the parameters are additive. For the interaction plots, when the lines between each point

of each color on each graph are parallel, there is no interaction. When the lines cross or

are not parallel, there is an interaction and the significance of the interaction is shown by

the difference in the slope of each line. Like the main effects, depending on the

determination of the importance of the interaction on the output, the combination of the

two parameters may or may not be significant in determining the sensitivity to the

system.

Figure 34. Interaction plots for field installations forward throw.

69

This evaluation was done for each diffuser output listed in Table 20, forward throw,

backward throw, sound and pressure, for all configurations in the L18 fractional factorial

test array. A predictive algorithm was developed using the main effects and interactions

data for each of the four output measures. In some cases, main effect and interaction

effect plots seem to indicate multiple effects occurring within the configurations,

however, the physics of the system was also used as a basis behind all determinations of

sensitivity. Thus if a reasonable physical explanation was not available for an effect, the

effect was not included in the model. Also, it was found that when certain main effects

and interactions were included in the predictive model, they did not improve the overall

accuracy of the prediction and where thus removed from the prediction equation. Each

output measure required an individual algorithm due to a different combination of

significant parameter effects on sensitivity. Sections 4.5.1 through 4.5.4 cover the

predictive equations for the four output measures.

4.5.1 Forward Throw Ratio (Tf)

Figure 33, Figure 34 and Table 22 were used to determine a predictive algorithm

model for forward throw. From Figure 33, the diffuser type graph shows a significant

sensitivity to the forward throw based on the type of diffuser used in the configuration.

Inlet diameter also is considered to have a large sensitivity effect on the forward throw.

These parameters are characteristics of the diffuser. Therefore these two values are

multipliers in the prediction equation. The other four parameters in Figure 33 are field

installation parameters. From the remaining four graphs of the main effects, duct height is

the only one that results in significant output variation and will be used as an additive

quantity to the multipliers. Duct type, damper, and approach angle have nominal effects

70

according to the graphs in Figure 33. Considering the interaction plots in Figure 34, it

was determined that the interaction between duct height and damper resulted in a change

from the sum effect of adding the individual main effects. Therefore, an interaction effect

is included. This quantity was used as an additive quantity just as the main effect of duct

height.

The predictive equation for the ratio of field installation forward throw to Standard 70

throw is:

FTRPrediction = MDT * MID * (ADCH + ADCH*D ) + 1

The parameters in this equation were quantified as follows based on Figure 33, Figure 34

and any anomalies discovered during experimentation.

Main Effect Parameters

For diffuser type;

Perforated Round: MDT = 0.8

Round: MDT = 1.0

Plaque: MDT = 0.6

Square: MDT = 1.0

Perforated Square: MDT = 1.0

Louvered: MDT = 1.0

71

For inlet diameter;

12-inch inlet diameter: MID = 1.2

8 or 10-inch inlet diameter: MID = 1.0

For duct height;

Duct height of 0: ADCH = 0.4

Duct height of 1.5: ADCH = 0.1

Duct height of 3: ADCH = 0

Interaction Parameters

For the interaction between duct height and damper;

Duct height of 0 and damper is perpendicular: ADCH*D = -0.2

Duct height of 0 and damper is not perpendicular: ADCH*D = 0

Duct height of 1.5 and damper is parallel: ADCH*D = -0.1

Duct height of 1.5 and damper is not parallel: ADCH*D = 0

Duct height of 3 and damper is 0 (no damper): ADCH*D = 0

Duct height of 3 and damper is a damper and damper type is not round opposed blade:

ADCH*D = 0.2

Concerning the last interaction shown, it was discovered that when the duct height was 3

diameters, the round opposed blade damper, compared to other dampers, caused less of

an increase in throw velocity in the forward direction. This equation gives a predicted

72

value that is compared to the actual value obtained for each configuration in Table 22 as

it varies from a reference value of one. An error is established between the two values

and then from this list of errors, a standard deviation is calculated to help determine if the

algorithm is accurate for all the configurations. The standard deviations in Table 22 show

that the current prediction algorithm is within 20 percent accuracy for the 18 test

configurations used in the field-testing experiment. To verify the prediction model, Table

23 shows random configurations that were installed and tested.

Table 22. Predicted values and corresponding errors for forward throw.

Experiment Tf

Error Experiment Tf

Error Experiment Tf

Error Run #

Predicted Tf 1200 fpm 800 fpm 500 fpm

1 1.40 1.53 -0.13 1.53 -0.13 1.53 -0.13 2 1.10 0.91 0.19 0.91 0.19 0.91 0.19 3 1.24 1.50 -0.26 1.42 -0.18 1.42 -0.18 4 1.12 1.04 0.08 1.04 0.08 1.04 0.08 5 1.00 0.69 0.31 0.69 0.31 0.69 0.31 6 1.00 1.01 -0.01 0.96 0.04 0.96 0.04 7 1.00 0.90 0.10 0.97 0.03 0.97 0.03 8 1.00 1.03 -0.03 1.03 -0.03 1.03 -0.03 9 1.19 1.16 0.03 1.16 0.03 1.12 0.07 10 1.20 1.47 -0.27 1.32 -0.12 1.28 -0.08 11 1.40 1.30 0.10 1.30 0.10 1.30 0.10 12 1.12 1.16 -0.04 1.25 -0.13 1.13 -0.01 13 1.10 1.04 0.06 1.04 0.06 1.04 0.06 14 1.20 1.13 0.07 1.26 -0.06 1.26 -0.06 15 1.48 1.62 -0.14 1.44 0.04 1.56 -0.08 16 1.20 0.89 0.31 0.89 0.31 0.98 0.22 17 1.40 1.32 0.08 1.32 0.08 1.32 0.08 18 1.12 1.38 -0.26 1.44 -0.32 1.37 -0.25 Average 0.01 Average 0.02 Average 0.02

Std

Deviation 0.17 Std Deviation 0.16 Std

Deviation 0.14

73

Table 23. Verification testing configurations.

Run #

Diffuser Type Inlet

Duct Height

Type Duct Damper

Damper Type

Approach Angle

1 Square 10 1.5 Rigid 0 RndSliding 0 2 Plaque 10 1.5 Rigid 0 RndOpBld 0 3 Plaque 10 1.5 Rigid Parallel RndSliding 0 4 Plaque 10 1.5 Rigid Parallel RndOpBld 0 5 Plaque 10 3 Rigid Parallel RndOpBld 0 6 Plaque 10 3 Rigid Parallel RndSliding 0 7 Plaque 10 0 Flex Parallel RndOpBld 0 8 Plaque 10 0 Flex 0 RndOpBld 0

It was determined that approach angle has no affect on the configuration accept as to

where the max throw exited the diffuser face and was left out of the verification testing.

The parameters in Table 23 were used as such to develop comparisons based on a few

important characteristics and to determine if the prediction model would hold true. The

configurations were tested exactly as before, but only for the highest noise condition to

test multiple configurations in less time. The results and prediction values are listed in

Table 24 along with the errors and standard deviations.

Table 24. Results of verification testing configurations for forward throw.

Experiment Tf Error Run #

Predicted Tf 1200 fpm

1 1.10 1.03 0.07 2 1.06 1.02 0.04 3 1.00 1.19 -0.19 4 1.00 0.90 0.10 5 1.00 0.88 0.12 6 1.12 1.15 -0.03 7 1.24 0.96 0.28 8 1.24 1.29 -0.05 Average 0.04

Std Deviation 0.14

74

Table 24 shows that the accuracy of the prediction model has a standard deviation well

below 20 percent for the verification configurations. Therefore, the algorithm should be

adequate in determining the performance of any ceiling diffuser installation forward

throw asymmetry to within acceptable limitations.

The same process was repeated for the remaining three output measures of the ceiling

diffuser installation configurations. The correction factors and prediction equations for

the other three output measures are listed in the following sections using the same process

used for the forward throw predictive algorithm.

4.5.2 Backward Throw Ratio (Tb)

The backward throw predictive algorithm was determined same as the forward throw.

A main effect plots were derived along with interactions for the same parameters as

forward throw. The main effects parameters, diffuser type and inlet diameter, have a large

sensitivity on the backward throw and are characteristics of the diffuser, while duct

height and damper have a smaller affect as they are field installation parameters. The

parameters that are characteristics of the diffuser are multipliers and the parameters that

are field installation related are additive quantities. The interactions between duct height

and damper, duct height and duct type, and duct height and inlet diameter resulted in a

change from the sum effect of adding the individual main effects leading to additive

quantities in the predictive algorithm.

The predictive equation for the ratio of field installation backward throw to Standard

70 throw is:

BTRPrediction = MDT * MID * (ADCH + AD + ADCH*D + ADCH*DCT + ADCH*ID ) + 1

75


For diffuser type;

Louvered: MDT = 1.2

Not Louvered: MDT = 1.0

For inlet diameter;




For duct height;

Duct height of 0: ADCH = -0.4

Duct height of 1.5: ADCH = -0.2


For Damper;

Damper is parallel: AD = 0.3

Damper is perpendicular: AD = 0.1

Damper is 0 (no damper): AD = 0

76



Duct height of 0 and damper is parallel: ADCH*D = 0.3

Duct height of 0 and damper is perpendicular: ADCH*D = 0.1


Duct height of 1.5 and damper is parallel: ADCH*D = 0.1

Duct height of 1.5 and damper is not parallel: ADCH*D = 0

Duct height of 3: ADCH*D = 0

For the interaction between duct height and duct type;

Duct height of 0 and duct type is flex: ADCH*DCT = -0.1

Duct height of 0 and duct type is not flex: ADCH*DCT = 0

Duct height of 1.5: ADCH*DCT = 0

Duct height of 3 and duct type is flex: ADCH*DCT = 0.1

Duct height of 3 and duct type is not flex: ADCH*DCT = 0

For the interaction between duct height and inlet diameter;

Duct height of 0 and 12-inch inlet diameter: ADCH*ID = 0

Duct height of 0 and 8 or 10-inch inlet diameter: ADCH*ID = -0.2

Duct height of 1.5 or 3: ADCH*ID = 0

77

From the given correction factors and the predictive equation, the following

experimental backward throw values were compared to the predicted values in Table 25

and gave the standard deviation of 0.16 or 16 percent for each flow rate velocity. With a

standard deviation below 20 percent, the accuracy of the predictive algorithm for

backward throw was confirmed. Verification testing was not done for backward throw

because it was found that forward throw, through verification, was accurate to within 20

percent and backward throw was given the same consideration.

Table 25. Predicted and corresponding values for backward throw.

Experiment Tb

Error Experiment Tb

Error Experiment Tb

Error Run

# Predicted

Tb 1200 fpm 800 fpm 500 fpm 1 0.46 0.45 0.01 0.49 -0.03 0.49 -0.03 2 0.90 1.05 -0.15 1.05 -0.15 1.05 -0.15 3 1.44 1.50 -0.06 1.50 -0.06 1.50 -0.06 4 0.64 0.47 0.17 0.47 0.17 0.50 0.14 5 1.20 1.37 -0.17 1.37 -0.17 1.37 -0.17 6 1.00 0.99 0.01 1.01 -0.01 1.01 -0.01 7 1.18 1.06 0.12 1.06 0.12 1.21 -0.03 8 1.00 0.99 0.01 0.99 0.01 0.99 0.01 9 0.67 0.81 -0.14 0.81 -0.14 0.59 0.08 10 1.18 1.12 0.06 1.07 0.11 0.99 0.19 11 1.00 0.96 0.04 0.96 0.04 0.96 0.04 12 0.78 0.83 -0.05 0.88 -0.10 0.93 -0.15 13 0.82 0.89 -0.07 0.89 -0.07 0.89 -0.07 14 1.10 1.12 -0.02 1.30 -0.20 1.30 -0.20 15 1.22 1.42 -0.20 1.29 -0.07 1.32 -0.10 16 1.32 0.84 0.48 0.84 0.48 0.94 0.38 17 0.16 0.16 0.00 0.14 0.02 0.13 0.03 18 0.87 0.68 0.19 0.70 0.17 0.56 0.31 Average 0.01 Average 0.01 Average 0.01

Std

Deviation 0.16 Std Deviation 0.16 Std

Deviation 0.16

78

4.5.3 Total Pressure Ratio (PTotal)

The total pressure ratio predictive algorithm was determined same as the forward

throw. A main effect plots was derived along with interactions for the same parameters as

forward throw. The main effects parameters diffuser type and inlet diameter have a large

sensitivity on the total pressure ratio and are characteristics of the diffuser, while duct

height, duct type, and damper have a smaller affect as they are field installation

parameters. The parameters that are characteristics of the diffuser are multipliers and the

parameters that are field installation related are additive quantities. The interaction

between duct height and damper resulted in a change from the sum effect of adding the

individual main effects leading to additive quantities in the predictive algorithm.

The predictive equation for the ratio of field installation pressure loss to Standard 70

pressure loss is:

TPRPrediction = MDT * MID * ( ADCH + ADCT + AD + ADCH*D ) + 1


For diffuser type;

Louvered: MDT = 1.4


Round: MDT = 1.0

Plaque: MDT = 1.2

Square: MDT = 1.2

79


For inlet diameter;



For duct height;




For duct type;

Duct type is flex: ADCT = 0.3

Duct type is mix: ADCT = 0.1

Duct type is rigid: ADCT = 0

For Damper;


Damper is not 0 and damper type is Round Sliding or Square Opposed Blade:

AD = 0.5

Damper is not 0 and damper type is Round Opposed Blade:

AD = 0.2

80




Duct height of 0 and damper is not 0: ADCH*D = -0.1

Duct height of 1.5 or 3: ADCH*D = 0

The given predictive equation for pressure and the correction factors developed the

predicted values given in Table 26 for each test configuration. Each was compared to the

experimental values to obtain the given error values. From the error values, the standard

deviations were determined and are less than 20 percent, giving an accurate predictive

algorithm.

In Table 27, the diffuser inlet configurations used for verification were tested and

gave the results listed in Table 28. With a standard deviation of 15 percent for all the

verification tests performed, the accuracy of the predictive pressure algorithm was

confirmed at 1200 fpm.

81

Table 26. Predicted and corresponding error values for pressure.

Experiment PTotal

Error Experiment PTotal


Error Run

# Predicted

PTotal 1200 fpm 800 fpm 500 fpm 1 1.72 1.78 -0.06 1.71 0.01 1.63 0.10 2 2.20 2.25 -0.05 2.14 0.06 2.17 0.03 3 1.72 1.85 -0.13 1.89 -0.17 1.70 0.02 4 2.32 2.40 -0.08 2.27 0.05 2.19 0.13 5 2.44 2.54 -0.10 2.35 0.09 2.17 0.27 6 1.14 1.19 -0.05 1.18 -0.04 1.16 -0.02 7 1.90 2.18 -0.28 2.09 -0.19 2.11 -0.21 8 1.30 1.23 0.07 1.22 0.08 1.24 0.06 9 1.78 1.73 0.05 1.78 0.00 1.75 0.03 10 1.60 1.46 0.14 1.65 -0.05 1.55 0.05 11 1.60 1.35 0.25 1.39 0.21 1.34 0.26 12 1.18 1.18 0.00 1.24 -0.06 1.25 -0.07 13 1.70 1.61 0.09 1.52 0.18 1.42 0.28 14 1.70 1.87 -0.17 1.91 -0.21 1.75 -0.05 15 1.66 1.50 0.16 1.48 0.18 1.41 0.25 16 2.12 2.18 -0.06 2.08 0.04 1.94 0.18 17 2.26 2.59 -0.33 2.56 -0.30 2.32 -0.06 18 1.76 1.84 -0.08 1.85 -0.10 1.72 0.03 Average -0.03 Average -0.01 Average 0.07

Std

Deviation 0.15 Std

Deviation 0.14 Std

Deviation 0.14


Run #

Diffuser Type Inlet

Duct Height

Type Duct Damper

Damper Type

Approach Angle

1 Square 10 1.5 Rigid 0 RndSliding 0 2 Plaque 10 1.5 Rigid 0 RndOpBld 0 3 Plaque 10 1.5 Rigid Parallel RndSliding 0 4 Plaque 10 1.5 Rigid Parallel RndOpBld 0 5 Plaque 10 3 Rigid Parallel RndOpBld 0 6 Plaque 10 3 Rigid Parallel RndSliding 0 7 Plaque 10 0 Flex Parallel RndOpBld 0 8 Plaque 10 0 Flex 0 RndOpBld 0 9 Plaque 10 0 Flex Parallel RndOpBld 0 10 Plaque 10 0 Flex 0 RndOpBld 0

82

Table 28. Results of verification testing for pressure.

Experiment PTotal

Error

Run # Predicted

PTotal 1200 fpm 1 1.48 1.2 0.28 2 1.48 1.18 0.30 3 2.08 1.63 0.45 4 1.72 1.33 0.39 5 1.48 1.41 0.07 6 1.84 1.64 0.20 7 2.20 1.92 0.28 8 2.08 1.516 0.56 9 2.20 2.103 0.10 10 2.08 1.791 0.29 Average 0.29

Std Deviation 0.15

4.5.4 Sound Level (NC) in Decibels

The sound (NC) level predictive algorithm was determined same as the forward

throw. A main effects plot was derived along with interactions for the same parameters as

forward throw. The main effects parameter diffuser type had a large sensitivity on the

sound level and was a characteristic of the diffuser, while duct height, duct type and

damper have a smaller affect, as they are field installation parameters. The parameters

that are characteristics of the diffuser are multipliers and the parameters that are field

installation related are additive quantities. The interactions between duct height and

damper resulted in a change from the sum effect of adding the individual main effects

leading to additive quantity in the predictive algorithm.

The predictive equation for the increase in NC level to Standard 70 NC level is:

NCPrediction = MDT * ( ADCH + ADCT + AD + ADCH*D )

83


For diffuser type;

Louvered: MDT = 1.1


Round: MDT = 1.3

Plaque: MDT = 1.3

Square: MDT = 1.1


For duct height;


Duct height of 1.5: ADCH = 3


For duct type;

Duct type is flex: ADCT = 1

Duct type is mix: ADCT = 1

Duct type is rigid: ADCT = 0

For Damper;


Damper is not 0 and damper type is round sliding: AD = 6.5

84

Damper is not 0 and damper type is square opposed blade: AD = 5.5

Damper is not 0 and damper type is round opposed blade: AD = 2



Duct height of 0 and Damper is 0 (no damper): ADCH*D = 0

Duct height of 0 and damper is not 0 and damper type is round sliding:

ADCH*D = -7

Duct height of 0 and damper is not 0 and damper type is square opposed blade:

ADCH*D = -5

Duct height of 0 and damper is not 0 and damper type is round opposed blade:

ADCH*D = -1

Duct height of 1.5 or 3: ADCH*D = 0

From the predictive equation for sound and the correction factors determined through

testing, Table 29 shows the error calculated between the predicted and experimental

sound data. The error and standard deviation for the sound level are not a percentage, but

measured in decibels. This means that the standard deviation at 1200 fpm varies 1.34 dB

for the initial test configurations. It was determined that a standard deviation of less than

3 dB would result in an accurate predictive algorithm. Again, due to the Standard 70

sound levels at 500 fpm for some diffusers being below the ambient of the UNLV throw

room, no analysis is presented for this data.

85

Table 29. Predicted values and corresponding errors for sound level.

Experiment NC

Error Experiment NC

Error Run

# Predicted

NC 1200 fpm 800 fpm 1 7.70 6.50 1.20 8.00 -0.30 2 11.55 11.50 0.05 11.50 0.05 3 9.35 8.50 0.85 8.50 0.85 4 9.75 10.50 -0.75 8.50 1.25 5 13.65 13.50 0.15 15.00 -1.35 6 1.30 5.50 -4.20 6.00 -4.70 7 6.65 9.50 -2.85 9.50 -2.85 8 1.40 2.00 -0.60 2.50 -1.10 9 5.25 5.50 -0.25 11.00 -5.75 10 3.00 2.50 0.50 2.50 0.50 11 3.00 3.00 0.00 4.00 -1.00 12 1.60 1.00 0.60 2.00 -0.40 13 5.20 7.00 -1.80 7.00 -1.80 14 9.75 9.50 0.25 10.00 -0.25 15 9.75 10.00 -0.25 9.50 0.25 16 8.25 8.00 0.25 5.50 2.75 17 8.80 8.50 0.30 12.50 -3.70 18 9.35 9.50 -0.15 11.00 -1.65 Average -0.37 Average -1.07

Std

Deviation 1.34 Std

Deviation 2.07

In Table 30, the verification configurations are listed and the resulting error calculations

are given in Table 31. The standard deviation was found to be less than 3 dB and

confirmed the accuracy of the predictive algorithm for sound at 1200 fpm.

86


Run #

Diffuser Type Inlet

Duct Height

Type Duct Damper

Damper Type

Approach Angle

1 Square 10 1.5 Rigid 0 RndSliding 0 2 Plaque 10 1.5 Rigid 0 RndOpBld 0 3 Plaque 10 1.5 Rigid Parallel RndSliding 0 4 Plaque 10 1.5 Rigid Parallel RndOpBld 0 5 Plaque 10 3 Rigid 0 RndOpBld 0 6 Plaque 10 3 Rigid Parallel RndOpBld 0 7 Plaque 10 3 Rigid Parallel RndSliding 0 8 Plaque 10 0 Flex Parallel RndOpBld 0 9 Plaque 10 0 Flex 0 RndOpBld 0 10

Plaque 10 0 Flex Parallel RndOpBld 0 11 Plaque 10 0 Flex 0 RndOpBld 0

Table 31. Results of verification testing for sound level.

Experiment Tf Error Run #

Predicted Tf 1200 fpm

1 3.30 4.80 -1.50 2 3.90 2.00 1.90 3 12.35 6.00 6.35 4 6.50 2.60 3.90 5 1.30 1.00 0.30 6 3.90 2.50 1.40 7 9.75 5.50 4.25 8

11.70 7.50 4.20 9 10.40 5.00 5.40 10 9.90 11.00 -1.10 11 8.80 7.50 1.30 Average 2.40 Std Deviation 2.60

4.5.5 Summary of Field Condition Testing Analysis

A summary of the standard deviations for the results and verifications of the field-

testing analysis are listed in Table 32. Note that verification testing was not done for

backward throw since the results for forward throw suggested good compliance with the

87

prediction equation. All verification tests were performed at the inlet velocity with the

highest noise level, 1200 fpm. The sound data has a standard deviation of decibels, not a

percentage like the other three outputs. Due to background sound levels, the diffuser

sound generation was not measurable for all field tests at the lowest inlet velocity

condition.

Table 32. Summary of standard deviations for field-testing analysis.

Noise Condition Verification 1200 fpm 800 fpm 500 fpm 1200 fpm

Forward Throw, Tf 0.17 0.16 0.14 0.14 Backward Throw, Tb 0.16 0.16 0.16 -- Sound, NC [dB] 1.34 2.07 -- 2.60 Pressure Ratio, PTotal 0.15 0.14 0.14 0.15

88

CHAPTER 5

CLOSE COUPLING AIR OUTLET PERFORMANCE

Diffuser performance measurements, where the installation used a short perpendicular

branch duct to the diffuser taken off a main duct, were conducted and compared to the

same diffuser performance from the Standard 70 characterization. That installation is

referred to as close coupling. The setup included an 11-inch by 18-inch rectangular main

duct that ran horizontally, approximately 16 feet in length and 1 foot above the ceiling. A

vertical duct section of varying heights and diameter, with a diffuser at one end was then

attached just beyond the midpoint of the rectangular main duct. The portion of the ceiling

that the main duct overshadowed was left open to allow for discharge air velocity

measurement of configurations where the shorter vertical duct heights put the diffuser

above the suspended ceiling. Due to the varying heights of the vertical branch and the

fixed position of the main duct in the ceiling, the ceiling height itself was varied to

always have the diffuser at the same height and to align the diffuser face with the array of

draft meters. The top six draft meters were the same height from the ceiling as that used

in the Standard 70 tests, but the total height of the array was reduced to 7 feet to

accommodate for the lower diffuser height. The variations included three vertical duct

heights going into the diffuser intake, three types of diffusers and use of a damper in the

intake. Discharge symmetry, pressure loss and noise generation were both affected by the

variations in installation. Invariably, the Standard 70 characterizations had the least

discharge asymmetry, lowest pressure loss and lowest noise generation. Increases in these

output measures were quantified by comparison to the output levels measured in a

89

Standard 70 installation. Installers and designers can use the quantitative results in this

report to predict the actual performance of a variety of close coupling installations.

5.1 Background

As the previous chapter showed, under typical field conditions, diffusers and outlets

can have throw, pressure loss and noise performance that is significantly different from

published performance obtained from ASHRAE Standard 70 testing. Particularly

applicable to this experiment are field installations, where a perpendicular branch duct is

tapped directly into a main duct with short or nearly no branch duct length. The

magnitude of the changes in the diffuser performance effects associated with these

installations can also depend on variations out of the control of the installer such as

variations in the main duct air velocity and duct static pressure.

5.2 Objective

The objective of this testing is to develop guidelines that will relate manufactures air

outlet cataloged data that have been obtained using ASHRAE Standard 70 to close

coupling installation applications. The testing collected data of diffusers obtained under a

collection of vertical branch inlet conditions and compared that data to base performance

data obtained from testing using ASHRAE Standard 70. From that comparison,

correction factors for diffuser throw and sound data were determined.

5.3 Methodology


90

1. Sound Power Level and Resulting Room NC.

2. Diffuser throw distance ratio of total throw in close coupling installation to throw

in a Standard 70 test taken in opposite directions enabling calculation of a throw

asymmetry ratio.

3. Ratio of main duct static pressure in field installation to total pressure in a Standard

70 test.


All tests for this project were run under steady-state conditions similar to the Standard

70 tests. Volume airflow corresponding to the required inlet velocity was set and allowed

to stabilize. The throw room measurement instrument configuration was the same as in

the Standard 70 testing except that the draft meter array was reduced to a vertical height

of 7 feet. The ceiling height was adjusted so that the diffuser discharge plane was 0.75

inches above the highest draft meter. Also, the same LabView virtual instrument was

used to monitor and record airflow, supply air temperature, room temperature, main duct

static pressure, draft meter array position and draft meter readings. Sound level

measurements were made with all systems off except for the air supply, boom

microphone and the monitoring and recording computer. Sound measurements were

taken with the air supply on and off to record the diffuser generated sound with airflow

and a background noise level without airflow.



diffuser discharge with, no pressure loss other than velocity pressure, no sound

generation and a discharge velocity pattern as intended by the diffuser design. Thus, any

91

variations from ideal would be variations in pressure measured in the duct upstream and

downstream of the diffuser inlet, increases in sound level measured in the discharge room

and unintended asymmetry in diffuser throw measured in the discharge room. Standard

70 measurements show variations from ideal due to flow restrictions in the diffuser,

turbulence induced in the flow due to diffuser structure and anomalies in discharge throw

due to the geometry of the diffuser design. The Standard 70 measurements obtained are

considered to be the closest to ideal with the diffusers tested. In this experiment, the

Standard 70 results are used as a standard while the test results are compared to that

standard. The experimental setup is diagrammed in Figure 35.

Figure 35. Experimental setup for close coupling laboratory testing.

92

Experimental output measures were obtained using the same or comparable methods

used for standard 70 testing and thus can be compared to Standard 70 measurements. The

pressure ring in the VAV unit at the end of the setup was used to determine the main duct

flow that bypassed the branch. Using the velocity pressure, the temperature of the air

traveling through the VAV, a correction value and a multiplier determined through

regression analysis of dynamic pressure, Pdyn = (1/2)ρv2, an equation was formulated to

calibrate the volume flow rate of air leaving the VAV unit, VAVFlow, and can be found in

Appendix A, section A.2.2. Airflow that went through the branch was calculated to be

total airflow minus the branch bypass airflow. These calculations were crosschecked

using a portable flow hood.

Sound measurements were recorded using the same method as the field installation

tests, using a rotating boom microphone pictured in Figure 11 in Chapter 3, where 1/3

octave band sound levels were recorded. Diffuser discharge velocities at incremental

distances from the diffuser were measured using the same array of draft meters at varying

heights in the room as used in the Standard 70 measurements. The array of draft meters

was scanned along the open ceiling perpendicular to the diffuser face pictured from the

top view in Figure 36. Due to the fixed position of the main duct and height variation of

the draft meters, the ceiling was varied in height to align the diffuser face with the top

draft meter.

93

Figure 36. Top view of close coupling scan pattern.

For this experiment, inlet main duct velocity was the noise condition. Noise

conditions, in product optimization, are conditions that are not controlled by the designer

but are set by the product user or result from external factors not controlled by either the

designer or the user. Airflow velocities used were 1150, 600, and 300 fpm in the

rectangular main duct. These velocities cover a large span of typical main duct flow

conditions including what is typically seen in actual installations.

Four parameters were used in the performance characterization experiment. They

were diffuser type, diffuser inlet diameter, vertical duct height directly above the diffuser

inlet and damper configuration in the diffuser inlet. The parameter and noise

configurations with levels are shown in Table 33. Three diffuser types were used in the

experiment. All these diffusers had fixed outlet sizes and differing inlet neck sizes. Thus

the three diffuser types had different free face area to duct area ratio as the inlet size

changed.

94

Table 33. Test parameters and noise conditions with different states.

Parameter State 1 State 2 State 3 1. Diffuser

Type Square Plaque Perforated Round Neck


3. Vertical Duct Height

0 Duct Diameters

1.5 Duct Diameters

3 Duct Diameters

4. Damper No Damper

Cross Damper X Damper

Noise Condition Low State Medium

State High State

1. Main Duct

Velocity 300 fpm 600 fpm 1150 fpm

The most economical (least run) fractional factorial array was used to set up the

experiment that could be used to extract the main effects of the test parameters and the

variation of those main effects due to the noise conditions. An L9 array was used. The

test array has 9 runs with four parameters at three levels. The L9 array has 8 degrees of

freedom (DOF), which is the total number of tests minus one. The total DOF needed for

the main effects are 8. The DOF needed for the main effects was determined by taking

the total number of levels of each parameter minus one and summing the result for all the

parameters (2 x 4). The array is shown in Table 34. Note that the branch duct connecting

the diffuser to the rectangular main duct was rigid for all installations. The damper used

was a round opposed blade damper and was always fully open. An example of a close

coupling configuration installation is shown in Figure 37. There were three sets of output

measures, one for each noise condition. The output array for one noise condition is shown

in Table 35.

95

Table 34. Test array for close coupling experiment.

Run Order

Run #

Diffuser Type Inlet

Duct Height Damper

2 1 Square 8 0 None 7 2 Square 10 1.5 X 6 3 Square 12 3 Cross 4 4 Plaque 8 1.5 Cross 5 5 Plaque 10 3 None 1 6 Plaque 12 0 X 8 7 Perf Rnd 8 3 X 9 8 Perf Rnd 10 0 Cross 3 9 Perf Rnd 12 1.5 None

Figure 37. Example close coupling configuration.

96

Table 35. Output array for the close coupling experiment.

Forward Throw

Asymmetry, Tf

Backward Throw

Asymmetry, Tb

Change in NC

Pressure Ratio

Run # Noise Level Noise Level Noise Level Noise Level 1 2 … 8 9

5.4 Results

Each diffuser output measure is a comparison of the close coupling configurations

against the corresponding data from the base Standard 70 tests. For example, the

Standard 70 velocity profile data in Figure 38 was used for comparison against the data

from the same diffuser type used in the close coupling tests in Figure 39 and Figure 40.

The asymmetry was determined by comparing the throw distance in zone 3 or zone 4 of

the zone plots from the close coupling configurations and the Standard 70 tests based on

the zone plot procedure described in Chapters 3 and 4.

To determine the zone throw distance at each recorded data point, the same approach

that was used for Standard 70 and field installation tests was done for the close coupling

installations. If the y2 value, reference Figure 28 from Chapter 4, of 0.2 was in zone 4,

the same procedure was followed from the field installation chapter using the beginning

of the zone 4 slope for x1 and y1, but the square root of the x2 value, calculated from the

comparison between close coupled and Standard 70 data, was taken to compensate for the

effect of the open ceiling.

97

Figure 38. Standard 70 throw data at 1200 fpm for 12” square diffuser.

Figure 39. Close coupling forward throw data at 1150 fpm (in main duct) for Run #3.

0 200 400 600 800

1000 1200 1400 1600 1800 2000

0 1 2 3 4 5 6

Velo

city

[ft/m

in]



0

200

400

600

800

1000

1200

1400

0 1 2 3 4 5 6

Velo

city

[ft/m

in]



98

Figure 40. Close coupling backward throw data at 1150 fpm (in main duct) for Run #3.

For example, Figure 41 and Figure 42 have an x2 value that occurs in zone 4 as the

y2 value of 0.2 was traced across through zone 4. This produced the throw ratio between

the close coupling data and the Standard 70 data. The change in NC level is a direct

difference between the NC of the close coupling configurations and the NC of the new

estimated Standard 70 data.

Using the volumetric flow rate values obtained from the VAVFlow equation and

comparison recordings using an Alnor ABT Balometer Capture Hood, approximated

volumetric flow rates were determined for each noise condition. To determine the new

Standard 70 data for 300, 600 and 1150 fpm, inlet neck area of the diffuser was used to

calculate the corresponding neck velocity at each flow rate. Using the original Standard

70 and published manufacturer data, estimations were made that represented new

Standard 70 data to compare to the close coupling output measure values. In some cases,

0

200

400

600

800

1000

0 1 2 3 4 5 6

Velo

city

[ft/m

in]



99

the NC levels at 300 fpm were below the room ambient sound level and have been left

out of the analysis. Table 36 shows the forward and backward throw ratios between the

close coupled and Standard 70 data at corresponding throw velocities. It also provides the

sound level differences between the close coupling configurations and Standard 70 as

well as the pressure ratio between the close coupling main duct static pressure and the

Standard 70 supply plenum total pressure. It was determined that for calculating diffuser

pressure loss within the vertical branch in this experiment, the static pressure in the

rectangular main duct was equivalent to the static pressure in the plenum used for

Standard 70 testing. The upstream pressure ring from the vertical branch measured static

pressure, but a small velocity pressure also existed within the duct. As the air moved

passed the opening for the vertical branch, velocity pressure was lost. At the second

pressure ring downstream from the vertical branch, static pressure increased to account

for the velocity pressure lose and any energy losses experienced between the two

pressure rings.

Figure 41. Run #3 forward throw zone plot data.

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


100

Figure 42. Run #3 backward throw zone plot data. Table 36. Diffuser output throw ratios and noise criteria level changes for close coupling compared to Standard 70 data.

5.4.1 Square Diffuser Performance Variation Effects when not Flush with the Ceiling

An important phenomenon was discovered during Run #1 when using the 8” square

diffuser without having a ceiling in line with the diffuser face. The testing of this diffuser

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


Forward Throw Asymmetry, Tf

Backward Throw Asymmetry, Tb

Change in Sound (NC) Level [dB]

Total Pressure Ratio, PTotal

Run # 300

fpm 600 fpm

1150 fpm

300 fpm

600 fpm

1150 fpm

300 fpm

600 fpm

1150 fpm

300 fpm

600 fpm

1150 fpm

1 0.958

1.0266

1.1661

0.8246

0.8246

0.8246

5 6 13 1.442

1.561

1.869 2 1.1

063 1.1063

1.1063

0.6884

0.9141

0.9141

9.5 11.5 17 1.649

1.874

1.898 3 1.0

266 1.1063

1.1063

0.9401

0.9401

0.9756

-- 11 19.5 1.210

1.396

1.439 4 0.9

608 1.0147

1.0987

1.032

1.032

1.1917

8.5 10 17 1.460

2.412

3.226 5 0.9

97 0.997

1.0987

0.8592

0.8592

1.0879

-- 7 19.5 1.791

1.891

2.576 6 0.9

728 1.0651

1.1716

0.9941

1.1716

0.9941

6.5 8.5 20.5 2.284

2.332

2.441 7 0.7

444 0.7444

0.7444

0.7958

0.7958

0.7958

6.5 11 13 1.775

1.911

1.677 8 0.9

542 1.0528

1.1255

0.8672

0.8672

0.8672

4 8.5 14.5 1.253

1.596

1.576 9 1.1

27 1.3412

1.574

0.8808

0.9868

0.9868

4 7 14 1.090

1.210

1.129

101

generated a zone plot that skipped past zones 2 and 3, directly into zone 4 as shown in

Figure 43.

Figure 43. Zone plot for forward throw of Run #1 without ceiling. The direct convergence into zone four means that the air being thrown perpendicular to

the diffuser side was nearly vertical directly after leaving the outlet diffuser face. This

behavior of the diffuser is uncommon and the test was repeated to ensure accuracy. To

determine if free air discharge was the issue, a ceiling flush with the diffuser over the

plane of measurement was installed as shown in Figure 44 and the test was repeated.

Figure 45 shows that with a ceiling, the diffuser performs more like a typical diffuser

than without a ceiling by spreading the flow along the ceiling and falling into zone 4 at a

much farther distance from the diffuser center.

0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


102

Figure 44. Styrofoam ceiling for Run #1 redo.

Figure 45. Zone plot for forward throw of Run #1 with ceiling.

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


103

5.5 Analysis

A statistical analysis was conducted to help better develop diffuser throw, pressure

loss and noise criteria prediction models for close coupling installations. The main effects

and interaction plots were determined based on the parameters listed in Table 33

individually for each output measure listed in Table 35.

To develop accurate and reliable prediction algorithms, as was done in Chapter 4 for

the field-testing analysis, the same algorithms were used for close coupling data that were

used for the field-testing data. Through acknowledgement of physics, certain parameters

were removed, while the parameters that were known to have an affect on the diffuser

performance, were kept in the predictive algorithm. The predictive algorithms for the

close coupling data most often turned out to have a worse standard deviation than the

field-testing configurations. One reason is likely the smaller testing array and fewer inlet

test parameters, thus fewer tests at each level of the test parameters. The predictive

algorithms are tabulated in sections 5.5.1 through 5.5.4.

5.5.1 Forward Throw Ratio (Tf)

An analysis of variance (ANOVA) was performed for the values of the forward throw

ratio against the four inlet parameters and was determined that three of the four

parameters had a significant affect on the diffuser performance. Relying on physics and

the predictive algorithm determined for the field-testing analysis, the following equation

to predict the forward throw ratio in a close coupling system was developed:

FTRPrediction = MDT * MID * ADCH + 1

104

The parameters in this equation were quantified as follows based on the main effects,

interaction plots and any anomalies discovered during experimentation.


For diffuser type;


Plaque: MDT = 0.6

Square: MDT = 1.0

For inlet diameter;



For duct height;




From the predictive algorithm for forward throw ratio, the predicted values for each

configuration tested were calculated in Table 37. The actual test results are listed in the

table as well in the experiment columns and the corresponding errors represent the

difference between the predicted and the experimental data. The standard deviations in

105

the table gave values that were slightly higher than the field-testing configurations and

the high state happens to be above the threshold of accuracy held at 20 percent.

Table 37. Predicted values and corresponding error for forward throw.

Experiment Tf

Error Experiment Tf

Error Experiment Tf

Error Run

# Predicted

Tf 1150 fpm 600 fpm 300 fpm 1 1.40 1.17 0.23 1.03 0.37 0.96 0.44 2 1.10 1.11 -0.01 1.11 -0.01 1.11 -0.01 3 1.00 1.11 -0.11 1.11 -0.11 1.03 -0.03 4 1.06 1.10 -0.04 1.01 0.05 0.96 0.10 5 1.00 1.10 -0.10 1.00 0.00 1.00 0.00 6 1.29 1.17 0.12 1.07 0.22 0.97 0.32 7 1.00 0.74 0.26 0.74 0.26 0.74 0.26 8 1.32 1.13 0.19 1.05 0.27 0.95 0.37 9 1.10 1.57 -0.48 1.34 -0.25 1.13 -0.03 Average 0.01 Average 0.09 Average 0.16

Std

Deviation 0.23 Std

Deviation 0.20 Std

Deviation 0.19

5.5.2 Backward Throw Ratio (Tb)

An ANOVA was performed for the values of the backward throw ratio against the

four inlet parameters and it was determined that two of the four parameters had a

significant affect on the diffuser performance. Relying on physics and the predictive

algorithm determined for the field-testing analysis, the following equation to predict the

backward throw ratio in a close coupling system was developed:

BTRPrediction = MDT * MID * ADCH + 1

106


For diffuser type;


Plaque: MDT = 0.6

Square: MDT = 1.0

For inlet diameter;




For duct height;

Duct height of 0: ADCH = -0.4

Duct height of 1.5: ADCH = -0.2


The predictive algorithm for backward throw ratio shows the calculated predicted

values for each configuration tested in Table 38. The standard deviations in the table gave

values that were slightly higher than the field-testing configurations but the threshold of

accuracy held at less-than-or-equal-to 20 percent was met.

107

Table 38. Predicted values and corresponding errors for backward throw.

Experiment Tb

Error Experiment Tb

Error Experiment Tb

Error Run #

Predicted Tb 1150 fpm 600 fpm 300 fpm

1 0.64 0.8246 -0.18 0.8246 -0.18 0.8246 -0.18 2 0.80 0.9141 -0.11 0.9141 -0.11 0.6884 0.11 3 1.00 0.9756 0.02 0.9401 0.06 0.9401 0.06 4 0.89 1.1917 -0.30 1.032 -0.14 1.032 -0.14 5 1.00 1.0879 -0.09 0.8592 0.14 0.8592 0.14 6 0.74 0.9941 -0.26 1.1716 -0.44 0.9941 -0.26 7 1.00 0.7958 0.20 0.7958 0.20 0.7958 0.20 8 0.68 0.8672 -0.19 0.8672 -0.19 0.8672 -0.19 9 0.82 0.9868 -0.16 0.9868 -0.16 0.8808 -0.06 Average -0.12 Average -0.09 Average -0.03

Std

Deviation 0.15 Std

Deviation 0.20 Std

Deviation 0.17

5.5.3 Total Pressure Ratio (PTotal)

An ANOVA was performed for the values of the total pressure ratio against the four

inlet parameters and it was determined that one of the four parameters had a significant

affect on the diffuser performance. Relying on physics and the predictive algorithm

determined for the field-testing analysis, the following equation to predict the total

pressure ratio in a close coupling system was developed:

TPRPrediction = MDT * MID * ( ADCH + AD ) + 1


For diffuser type;


108

Plaque: MDT = 1.2

Square: MDT = 1.2

For inlet diameter;



For duct height;




For Damper;


Damper is not 0: AD = 0.2

The predictive algorithm for total pressure ratio, main duct upstream static pressure

versus Standard 70 supply plenum total pressure, shows the calculated predicted values

for each configuration tested in Table 39. The standard deviations in the table gave values

that were much higher than the field-testing configurations and the threshold of accuracy

held at less-than-or-equal-to 20 percent was not met.

109

Table 39. Predicted values and corresponding errors for total pressure ratio.

Experiment PTotal



Error Run #

Predicted PTotal

1150 fpm 600 fpm 300 fpm 1 1.72 1.869 -0.15 1.561 0.16 1.442 0.28 2 1.72 1.898 -0.18 1.874 -0.15 1.649 0.07 3 1.29 1.439 -0.15 1.396 -0.11 1.210 0.08 4 1.72 3.226 -1.51 2.412 -0.69 1.460 0.26 5 1.24 2.576 -1.34 1.891 -0.65 1.791 -0.55 6 1.58 2.441 -0.86 2.332 -0.76 2.284 -0.71 7 1.40 1.677 -0.28 1.911 -0.51 1.775 -0.37 8 1.80 1.576 0.22 1.596 0.20 1.253 0.55 9 1.24 1.129 0.11 1.210 0.03 1.090 0.15 Average -0.46 Average -0.28 Average -0.03

Std Deviation 0.62 Std

Deviation 0.38 Std Deviation 0.42

5.5.4 Sound Level (NC) in Decibels

The close coupling test system was a substantially new room configuration and ran

the airflow at higher levels than the field-testing configuration. Therefore the close

coupling test room configuration was evaluated for background noise levels at all test

conditions. It was determined that at a main duct airflow rate of 1150 fpm, the supply flex

duct at 14 inches in diameter, generated breakout sound levels significantly above the

throw room ambient and near the test run recorded sound level data. Therefore, the test

run sound data for the 1150 fpm runs was not used in the analysis.

Initial analysis of the sound data showed that a test condition other than the four test

parameters was affecting the recorded sound levels. Figure 46 shows the correlation

between main duct velocity and change in NC levels. The points on the graph show a

large variance at each main duct velocity due to different levels of the four test

parameters, but the average NC level shows a significant increase with an increase in the

110

main duct velocity. The linear regression does not cross through the origin because the

damper has a minimum of approximately 3 dB affect on the output, which if the

regression line was extended towards the origin, it crosses at a delta NC value of about 3

dB.

Figure 46. Comparison between change in sound [dB] and main duct velocity.

The following regression equation was developed to calculate an estimated increase

in NC level based on the main duct velocity with 31.6 percent of the variance accounted

for from the coefficient of determination.

ΔNC [dB] = 0.009 * ( VelocityMD ) + 3.63

y = 0.0089x + 3.627 R² = 0.31635

0

2

4

6

8

10

12

14

0 100 200 300 400 500 600 700

Del

ta N

C [d

B]

Velocity in Main Duct [fpm]

Correlation Points Linear(Correlation Points)

111

Using the statistical software Minitab, a Pearson Product Moment Correlation was

used to prove that the correlation between the main duct velocity and the difference in

sound level was indeed accurate. The Pearson Correlation converts the ratio data into z-

scores so that each set of data are on the same scale. It is a linear dependant function that

compares two variables, X and Y, and uses hypothesis testing to determine the

probability that X and Y result in Pearson Correlation of zero. The closer the P-value is to

zero, the stronger the rejection is of the null hypothesis, ρ = 0. [13,14]

€

ρ =

xi − x ( ) yi − y ( )i=1

n

∑n −1( )sxsy

where:

€

x = sample mean for first variable

sx = standard deviation for first variable

€

y = sample mean for the second variable

sy = standard deviation for the second variable

n = column length

Note that the data columns must be the same length.

The closer the Pearson Correlation coefficient to one or negative one, the more the

variables are correlated and since the values in comparison are both increasing, the

correlation approaches one. Also, if the probability value (P-value) is less than the

significance level of 0.05, the significance of the correlation is proven strong on the 95

percent confidence interval. If the P-value were greater than the significance level of

112

0.05, the two variables would have an unlikely correlation. [13,14] With the sound data

adjusted based on the damper configuration, the Pearson Correlation was determined to

be 0.562 with a Probability value (P-value) of 0.023. Therefore, the correlation is

significant and the two sets of data are not random.

Based on the velocity having a significant affect on the diffuser performance, the

velocity in the main duct was accounted for by taking the velocity times the slope of the

regression line (0.009*VelocityMD) described previously and subtracting that value from

the NC levels in Table 36. This resulted in NC levels listed in Table 40.

Table 40. Adjusted NC levels that account for damper and not main duct velocity.

Change in Sound (NC) Level [dB]

Run # 300 fpm 600 fpm 1 2.30 0.60 2 6.80 6.10 3 -- 5.60 4 5.80 4.60 5 -- 1.60 6 3.80 3.10 7 3.80 5.60 8 1.30 3.10 9 1.30 1.60

Performing an ANOVA on the adjusted NC levels against all four of the inlet

parameters, the one parameter that showed significance with a P-value below 0.05 was

the damper at 0.002. The damper was determined independent of the other three

parameters. With this in mind, the main effects plot suggested a 3 dB increase when a

damper was present in the installation between 300 fpm and 600 fpm. The following

113

predictive algorithm for NC calculation in a close coupling system was developed to

reflect the affects of both main duct velocity and damper installation.

NCPrediction = AD + AMD * VMD


For damper;


Damper is not 0: AD = 3

For main duct velocity;

Additive correction coefficient for main duct velocity in ΔNC/fpm:

AMD = 0.009

Main duct velocity in fpm: VMD

The predictive algorithm for sound (NC) level, the difference between the

experimental and the Standard 70 data in decibels, shows the calculated predicted values

for each configuration tested in Table 41. The standard deviations in the table gave values

that were slightly lower than the field-testing configurations and the threshold of

accuracy held at less than 3 dB was satisfied. Also note that the Standard 70 data for the

high state of 1150 fpm was unreliable, as stated previously, because it experienced added

sound penetration from the supply duct to the background NC level and was omitted.

114

Table 41. Predicted values in dB and corresponding errors for sound (NC) level.

Predicted NC [dB]

Experiment NC [dB]

Error [dB]

Predicted NC [dB]

Experiment NC [dB]

Error [dB]

300 fpm 600 fpm 2.70 5.0 -2.3 5.40 6 -0.6 5.70 9.5 -3.8 8.40 11.5 -3.1 5.70 -- -- 8.40 11 -2.6 5.70 8.5 -2.8 8.40 10 -1.6 2.70 -- -- 5.40 7 -1.6 5.70 6.5 -0.8 8.40 8.5 -0.1 5.70 6.5 -0.8 8.40 11 -2.6 5.70 4.0 1.7 8.40 8.5 -0.1 2.70 4.0 -1.3 5.40 7 -1.6

Average -1.44 Average -1.54

Std

Deviation 1.77 Std

Deviation 1.10

5.5.5 Summary of Results for Close Coupling Testing

A summary of the results for the close coupling testing configurations is shown Table

42. Even though prediction equations have been established, it may be conducive to

conduct further experimentation, including more testing configurations, to achieve a

better determination of which inlet parameters have even more of a significant affect on

diffuser performance.

Table 42. Summary of standard deviations for close coupling testing.

Noise Condition 1150 fpm 600 fpm 300 fpm

Forward Throw, Tf 0.23 0.20 0.19 Backward Throw, Tb 0.15 0.20 0.17 Sound, NC [dB] -- 1.10 1.77 Pressure Ratio, PTotal 0.62 0.38 0.42

115

CHAPTER 6

CONCLUSIONS AND RECOMMENDATIONS

This investigation obtained quantitative diffuser throw, back-pressure and noise

generation data on the effects of field installation variations compared to published

Standard 70 data. The data, when presented in an easily usable form, can aid designers

and installers in their efforts to produce energy efficient, comfortable, quiet and effective

air distribution systems. All four phases of the investigation, plenum design, baseline

testing, field installation testing and close coupling testing produced such quantitative

data.

6.1 Diffuser Supply Plenum Design

The test results analysis from the plenum optimization experiment identified a

method of introducing air into the plenum and a design and position of the flow

equalization device that produced diffuser inlet velocity conditions that had variation well

below the requirement for Standard 70 testing. Perhaps more importantly, this

investigation resulted in guidance on plenum design that can be used for future plenum

builds to optimize the airflow into the diffuser inlet using robust design. The robust

design method was a good tool for economically deriving main effects of the system

parameters and showed the necessity to consider the physics and interactions in order to

perform verification tests and identify accuracy. Through verification testing, it was

determined that the ducted method and plenum method of testing suggested by Standard

70, gave similar results for throw, sound and pressure data when performed in the UNLV

throw room. The final results show that variation significantly lower than the Standard 70

116

specification can be achieved with proper plenum design. The results also indicate that a

minimum straight inlet duct of 7-10 inches be specified in the design regardless of low

airflow variation without the added duct length.

This plenum design experiment helped to develop testing methods that were later

used in the Standard 70 and field condition testing. The use of orthogonal test arrays for

determining the experiment test run conditions proved to be efficient and effective for

output analysis. Similar arrays used in the field condition testing analysis will allow

practical testing with multiple parameters.

6.2 Standard 70 Characterization

Following the guidelines in Standard 70, a baseline data set for comparison to real

field installations was developed. The data conforms to the linear regression described by

the zone plot formulation. The zone plot method produced characterizations of different

diffusers with differing inlet neck sizes and flow rates. The method also showed that

while similar diffusers from different manufacturers often perform nearly identical, that is

not always the case. Sufficient data was obtained to perform comparisons between

baseline and field installation performance. From this comparison, a set of calculated

algorithms were developed to determine the effect each physical parameter has on the

field installations. The Standard 70 plenum method was shown to be an accurate and

economical method to test for ideal diffuser performance.

117

6.3 Field Installations

Prediction algorithms for the performance difference ratio between field installations

and Standard 70 data were obtained. These algorithms are adequate for determining how

a ceiling diffuser installation is going to perform based on inlet configuration

characteristics. The quantitative data reflects averaged results across different

manufacturers and different airflows while taking into account inlet condition parameters

such as duct type, approach angle, damper type, damper orientation and vertical duct

height attached to the diffuser inlet. Several design conditions cause variation in throw

asymmetry, pressure and sound generation of the test diffuser. Also, the different types of

diffusers and different inlet to outlet area ratios affected the performance sensitivity of

the installation to variations in installation.

For throw asymmetry, duct height was the most important installation parameter. No

vertical duct leading into the diffuser resulted in significant asymmetry. A damper in the

inlet, depending on damper blade orientation, could reduce that asymmetry. The larger

inlet diameter (higher inlet to outlet area ratio) was more sensitive to installation, causing

more throw asymmetry than smaller diameters.

Both a damper in the diffuser inlet and the duct configuration that led into the diffuser

inlet affected pressure drop of the system. A damper at the diffuser inlet increased

pressure drop. Flex duct caused a higher pressure drop than rigid duct. When the duct

was connected directly to the diffuser inlet with no vertical duct height, pressure drop

was increased. The louvered diffuser was most sensitive to installation variation while the

modular core perforated square diffuser was the least sensitive. A larger inlet diameter

118

(higher inlet to outlet area ratio) reduced sensitivity to installation, causeing performance

variation.

Diffuser generated sound (NC level) was most affected by vertical duct height

above the inlet with no vertical duct resulting in a 7 dB increase in NC level. Damper

installation caused a minimum 2 dB increase and up to a 7 dB increase. Duct height and

damper effects are not always additive, such that the combination of no vertical duct and

a damper generate less noise than the sum of both effects. Plaque and round diffusers had

the highest sound performance sensitivity to installation variation while the modular core

perforated square had the least.

6.4 Close Coupling

In a close coupling installation with all rigid duct, the effect of diffuser type, branch

length and damper were not predictable using the exact same field installation formulas.

For close coupling, the main duct velocity and damper installation were the significant

factors in the determination of sound generation in a close coupling field installation. As

the velocity increases, the ratio of sound generation to the Standard 70 sound levels

increases. At slower main duct velocities, it was determined that for every 100 fpm, add

one decibel to the sound level prediction. Because the main duct velocity had such an

effect, a calculation was made to remove this affect from the sound data, so that only the

installation of a damper, which was independent of the other three parameters, had an

effect on the diffuser performance. The resulting data was used to develop the predictive

algorithm. Quantitative data for predicting the noise generation were determined. It was

119

also found that to achieve the proper flow velocity profile for any diffuser in the close

coupling position, a ceiling should be flush with the diffuser outlet face.

6.5 Recommendations for Designers and Installers

The algorithms developed in this investigation should be used to develop more

efficient, effective, comfortable and quiet air distribution systems. The formulas can also

be simplified into an easy to use table of corrections for most of the common installation

variations. Anyone building a laboratory test system using a plenum air supply will find

the plenum design results to be useful. Consideration should be given to have Standard

70 diffuser inlet airspeed variation specification given in terms of a standard deviation

from the mean, which more accurately reflects the quality of the inlet flow. Also Standard

70 should specify a minimum vertical duct height from the diffuser inlet to the inlet cone

regardless of flow uniformity.

120

APPENDIX A

TEST FACILITY

The following information describes the laboratory instrumentation, equipment and

the software used to manipulate each. Certain instruments were project specific and thus

were labeled as such. Most instruments and equipment pieces are physically apart of the

laboratory and are usable for much more than what was needed for this project.

A.1 Throw Room Description

The Center of Mechanical & Environmental Technology (CMEST) throw room is a

unique, state-of-the-art, automated testing facility. It can be used to measure the quality,

efficiency, and effectiveness of different heating, ventilating, and air conditioning

(HVAC) system components and configurations, and their related effects on building

occupants. The throw room is 30 ft (9.1 m) long by 20 ft (6.1 m) wide. The height of the

room ceiling, measured from the floor, can be varied from 7 ft (2.1 m) to 11 ft (3.3 m).

The throw room is equipped with both a traditional air distribution (CAD) system and

UFAD system. It can be rapidly reconfigured from CAD to UFAD systems with different

diffuser, grill, and load configurations. It can be easily set up as an office space with

partitions, a moderate size meeting room, a classroom, or a hotel suite. The system can

also be easily modified to add a displacement ventilation system by putting a

displacement terminal unit against the north wall where the vertical shaft for UFAD

systems is located. The side view of the throw room is shown in Figure 47. The figure

shows the supply air fan chamber and ASME nozzle chamber used to measure and

monitor the supply airflow rate to the throw room, the UFAD plenum, the interior room

121

space, the adjustable height room ceiling, the ceiling plenum, CAD and UFAD system

duct and the vertical shaft for UFAD system. The throw room has precision measurement

capabilities including [15]:

1. Airflow and room air distribution characteristics of grills, registers, diffusers, and

other types of room ventilation devices.

2. Airspeed, temperature, humidity and contaminant concentrations at precise points

throughout the entire room.

3. Temperatures at multiple wall, floor, ceiling, under-floor, and above ceiling

airspace locations.

4. Airflow, temperature, humidity, contaminant concentrations and pressure at

various locations in the supply and return ductwork and plenums.

5. Energy inputs from interior room, room lighting, and simulated exterior solar

loads and energy consumption of the HVAC system.

Its precision control capabilities include [15]:

1. Perimeter heating and cooling loads through the control of all room surface

temperatures by zone. Each room surface (includes walls, floor below the raised

floor, and ceiling above the dropped ceiling) is divided into one to three zones.

The surface temperature of each zone can be independently controlled. Each zone

can be heated or cooled independently as required to maintain the desired wall

temperature.

2. Interior room heating and cooling loads. Load simulators for humans and

machines can be placed throughout the room. Their individual and total energy

122

consumption can be monitored and controlled.

3. Supply airflow at 2,000 CFM BI centrifugal fan supplies air to the room, in a

closed loop or ventilated mode.

4. Supply air temperature is digitally controlled heat pump supplies air at preset

supply air temperatures to the room, in a closed loop or ventilated mode.

5. Precision automated instrument positioning from a computer-controlled, two-axis

(two translational axes) traversing system can hold multiple instruments for

measuring airspeed, temperature, sound, contaminant concentration and humidity

at any point in the room. The control system automatically moves the instruments

to the test positions while capturing and recording the measurements at those

positions. The sensors on the traversing mechanism are fixed.

6. Automatic monitoring of test conditions, both prior to start and during testing.

Operator warnings are sent if test and measurement conditions drift out of

predefined limits. Tests can be paused, resumed, or halted at any time.

123

Figure 47. Side view of the UNLV throw room [15].

A.2 Throw Room Data Acquisition Instrumentation

The following information describes the instrumentation used to control and verify

the thermal conditions in the UNLV throw room. A central computer is used to control

the instrumentation, data acquisition and motion control systems using a custom written

LabView program. This software will simultaneously monitor the room conditions and

instrument performance. A test could be stopped, continued or restarted at any point in

time with this software. Figure 48 and Figure 49 show the interfaces used by LabView

for monitoring the sensors and supply duct parameters in the UNLV throw room. Table

43 shows the instrumentation used by the National Instruments LabView program to

record supplemental data at a user set time interval of 4 seconds per sample. Table 44

20

capturing and recording the measurements at those positions. 5) Automatic monitoring of test conditions, both prior to start and during testing: Operator warnings are sent if test and measurement conditions drift out of predefined limits. Tests can be paused, resumed, or halted at any time. The throw room is 30 ft (9.1 m) long by 20 ft (6.1 m) wide. The height of the room ceiling, measured from the floor, can be varied from 7 ft (2.1 m) to 11 ft (3.3 m). The throw room is equipped with both a traditional air distribution (CAD) system and UFAD system. It can be rapidly reconfigured from CAD to UFAD systems with different diffuser, grill, and load configurations. It can be easily set up as an office space with partitions, a moderate size meeting room, a classroom, or a hotel suite. The system can also be easily modified to add a displacement ventilation system by putting a displacement terminal unit against the north wall where the vertical shaft for UFAD systems is located. The side view of the throw room is shown in Figure 3. The figure shows the supply air fan chamber and ASME nozzle chamber used to measure and monitor the supply airflow rate to the throw room, the UFAD plenum, the interior room space, the adjustable height room ceiling, the ceiling plenum, CAD and UFAD system duct and the vertical shaft for UFAD systems. Figure 4 shows iso-velocity contours measured in the throw room for a linear slot diffuser. The contours were obtained from airflow velocity data obtained using the four-axis computer-controlled traversing system.

Figure 3: Side View of the UNLV throw room

124

shows the equipment used to control the supply duct air flow characteristics and measure

supplemental data with and without LabView. A separate interface was used to control

the motion of the throw room traversing mechanism that had the throw velocity sensors

attached given in Figure 50. To set up the path that the sensors followed, another

interface was called upon through the motion control interface in Figure 51, where the

red box signified the diffuser dimensions. This interface could produce alternate path

types other than linear, such as, angled, raster and three sided to accommodate any

situation. All the instrumentation and equipment worked together to produce isothermal

conditions within the throw room during periods of testing.

Figure 48. Labview main interface for monitoring system conditions.

125

Figure 49. Interface for monitoring throw room surface temperatures.

Figure 50. Throw room LabView motion controller interface.

126

Figure 51. Throw velocity sensor path setup interface.

127

Table 43. Throw room instrumentation used along with NI LabView.

Sym

bol

No.

Suff

ix

Measured Parameter

Instrument Name and

Model

Qua

ntity

Location

Sampling, Storage

and Display

Rate, [Hz] Accuracy

TV 1 - 17 t,v

Air Temperature

(oC) and Velocity

(m/s)

Sensor Technology

HT-400 17

Traversing Mechanism

pole 0.5

Temperature: ±2 oC

Velocity: 0.05 to 1m/s, ±1% of reading; 1 to 5m/s, ±3% of

reading

TH 1 - 2

Air Temperature

(oC) and Humidity (%)

Omega HX93 AV 2

Room (TH1) and Ceiling

Plenum (TH2) 0.5

Temperature: ±0.5 oC Relative

Humidity: ±2.5%

TH 3 - 5 d

Air Temperature

(oC) and Humidity (%)

Omega HX93 AV-

D, Duct Mounted

3 Supply (TH3d-4d) and Return

Ducts 0.5

Temperature: ±0.5 oC Relative

Humidity: ±2.5%

TC 1 t Air

Temperature (oC)

Omega Thermal

Couples, T type with

24 in. probe

1 Nozzle Chamber 0.5 Temperature:

±0.5 oC

TC 2 - 25

Air Temperature

(oC)

Omega Thermal

Couples, K type with

12 in. probe

24

Ceiling Supply Diffusers (TC2-

5), Floor Supply

Diffusers (TC6-13), Floor

Plenum (TC14-22), Ceiling

Plenum (TC23-25)

0.5 Temperature: ±1.1 oC

T4 1 - 97

Wall Temperature

(oC)

Omega Thermal Couples,

SA1K

97 Walls, Ceiling, Floor 0.5 Temperature:

±1.1 oC

P 1 - 3 d

Static or Differential Pressure (in.

H2O)

Omega PX277-05D5V

3

Nozzle Chamber (P1d-

2d), Supply Plenum (P3d)

0.5 Pressure: ±1% of full scale

P 4 - 15

Static Pressure (in.

H2O)

Omega PX277-01D5V

12 Close Coupled Main/Supply Duct (P4,P6),

0.5 Pressure: ±1% of full scale

128

Return Duct and Floor

Plenum (P5,P7-P15)

TF 1 - 2

Air Temperature

(oC) and Flow Rate

(cfm)

EBTRON GTx116-Pc 2

Supply (TF1) and Return

(TF2) Ducts 0.5

Temperature: ±0.075 oC

Airflow Rate: ±2% to 3% of

reading Table 44. Testing equipment used in part with throw room.

Instrument Name and

Model Measured Parameter Location Accuracy Note

TSI VelociCalc Plus Meter

8386A

Velocity (m/s),

Temperature (oC), and Relative

Humidity (%)

Handheld Device

Velocity: ±3% of reading

Temperature: ±0.25oC Relative

Humidity: ±3%

Used for verification and plenum optimization

data

TSI Alnor LOFLO

Balometer 6200

Airflow Rate (cfm)

Handheld Device

Flow Rate: ±(3%+5cfm)

Range: 10 to 500 cfm

TSI AccuBalance

Airflow Capture Hood

8371

Airflow Rate (cfm)

Handheld Device

Flow Rate: ±(5%+5cfm)

Range: 30 to 2000 cfm

Bruel & Kjaer Type 4942

Sound (dBA)

Boom Microphone

NC: 0.2 dB at 95% CI

Capture Sound Data in Throw

Room

129

Table 45. Systems used in part with the throw room.

System Name and Model

Measured Parameter Location Accuracy Note

Greenheck Centrifugal Fan

15-BISW-I

Airflow Rate (cfm)

Blower Room

Produces Airflow for

Throw Room Max 2674 rpm

3-Phase Lincoln Motor

(Leeson) SSF4P5T61

Frequency (Hz)

Blower Room

Fan Motor, 5 HP, 60 Hz,

1745 rpm

PDL Electronics

LTD XTRAVERT

X712

Fan Frequency

(Hz) Lab Room 97% Electrical

Efficiency

Controls Motor That

Controls Fan

White Rodgers IF93-380

Temperature (oC) Lab Room

Thermostat w/ Setpoint Range (7-

37oC)

Daikin Unit VRV II

REYQ96MTJU

Temperature (oC) Outside

Cool/Heat Walls of Throw Room

Bryant Heat pump

213RNA036-D

Temperature (oC) Outside

Cool/Heat Supply Air 1.5-5 Tons

Titus VAV box

Pressure (in. H2O),

Temperature (oC)

Throw Room Variable Air

Volume

A.2.1 Calibration of the Ebtron flow measurement unit

Also note that the ASME approved nozzle chamber used complicated calculations

with the LabView interface to provide accurate throw into the system that was measured

130

by the Ebtron unit installed at the supply duct. A test was run to check the calibration of

the calculations, see Figure 52, and it was found that the setup was accurate up to a flow

rate of 1,000 cubic feet per minute, where soon thereafter, the data showed increasing

degradation in accuracy of the nozzle calculations with increasing flow rate. The

configuration of the nozzle chamber was changed for each range of flow according to it’s

manufacturer calibration. The points that represent F1 are the points that were set by the

tester. The points that represent Q_R are the calibrated values measured by the Ebtron

unit after the air is pumped through the nozzle chamber.

Figure 52. Calibrated nozzle chamber versus Ebtron unit.

0 100 200 300 400 500 600 700 800 900

1000 1100 1200 1300 1400 1500 1600 1700 1800

Cal

ibra

ted

Flow

Rat

e Fr

om N

ozzl

es [c

fm]

User Set Flow Rate From Nozzles [cfm]

Flow Rate Test at 50 cfm Increments

Q_R F1

131

A.2.2 Calibration of the VAV unit

The variable air volume (VAV) unit was calibrated as well before conduction the

close coupled tests to better determine the amount of air leaving the vertical branch. Like

the Ebtron unit, the VAV had a pressure ring that was used to determine the flow of air at

a given point in time. The damper built into the VAV was fully open and the supply air,

determined by the Ebtron, was increased at increments to determine the pressure increase

read by the pressure ring inside the VAV. The value for the flow rate of F1 shown in

Figure 53 was squared to linearize the correlation with pressure of P3d. At more than

99.5 percent of the variance accounted for, the two parameters are highly correlated.

Figure 53. VAV unit calibration.

Taking the total pressure of the duct system and subtracting the static pressure

measured at the VAV, the velocity pressure was determined. Velocity pressure is

y = 1E-07x - 0.0001 R² = 0.99527

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

Pres

sure

P3d

[in-

H2O

]

Flow Rate F12 [cfm]2

F1 vs P3d

Linear(F1 vs P3d)

132

calculated from measurement of the total airflow and knowledge of the main duct cross-

sectional area. Using the velocity pressure, the temperature of the air traveling through

the VAV, a correction value and a multiplier determined through regression analysis of

dynamic pressure, the following equation was used to determine the volume flow rate of

air leaving the VAV unit, where TVAV is the temperature of air inside the VAV unit and

PVAV is the pressure inside the VAV unit.

VAVFlow = 2740 * {[( TVAV + 273 ) / 297 ](1/2) } * ( PVAV )(1/2) - 29

Using the above equation and the known value of the incoming flow from the Ebtron

flow measuring unit, the difference between the two yields the flow exiting the diffuser

attached to the branch on the rectangular main duct in the close coupling design shown in

Figure 35.

A.3 Project Specific Data Acquisition Instrumentation

The following instrumentation is specific to the experimentation done for this

research. The plenum was built in-house per ASHRAE Standard 70-2006 and meets the

minimum requirements from this standard. The boom microphone was used to calculate

the room NC levels during testing.

A.3.1 Optimized supply plenum leak test

The data points shown in Figure 54 represent a leak test that was conducted after the

finalized construction of the optimized plenum before any Standard 70 baseline tests

were performed. The supply air inlet conditions were kept at the optimal settings and the

133

opening for the diffuser installation was sealed. Increasing the airflow into the plenum

caused the pressure to increase within the plenum. These two parameters, flow and

pressure, were correlated with one another to determine the leakage from the plenum. At

nearly 99 percent of the variance accounted for in the correlation, the leakage from the

supply plenum is very small.

Figure 54. Plenum pressure test. A.3.2 Boom microphone throw room calibration

The boom microphone was setup to capture an averaged sound level within the throw

room by recording points at different locations in the reverberant field. The boom

microphone, pictured in Figure 11, rotated 180 degrees and recordings were made every

five seconds as the boom traversed along it’s path. All recorded points were averaged

over a minimum of 60 seconds. NC contour curves, shown in Figure 55, were used to

determine the point of tangency with the highest-ranking NC curve. [6,8]

y = 55.354x - 8.1137 R² = 0.98546

0

5

10

15

20

25

30

0.1 0.2 0.3 0.4 0.5 0.6 0.7

Flow

F1

[cfm

]

Pressure P3d [in-H2O]

P-Test Linear(P-Test)

134

Figure 55. Published NC contour curves. This set the NC rating for the current configuration of the UNLV throw room and for

each 1/1 octave band center frequencies, given below, an equation was formulated

specific to the UNLV throw room. Another LabView interface was used to accomplish

this and is shown in Figure 56.

Boom microphone NC calculations:

125 Hz [dB] = X1 - X2 - 10 - 64 ·108 + 50

250 Hz [dB] = X1 - X2 - 10 - 58 · 108 + 50

500 Hz [dB] = X1 - X2 - 10 - 54 · 109 + 50

1000 Hz [dB] = X1 - X2 - 10 - 51 + 50

2000 Hz [dB] = X1 - X2 - 10 - 49 + 50

4000 Hz [dB] = X1 - X2 - 10 - 48 + 50

0

10

20

30

40

50

60

70

80

125 250 500 1000 2000 4000

Soun

d Pr

essu

re L

evel

s [dB

]

Octave Band Center Frequency [Hz]

NC-15 NC-20 NC-25 NC-30 NC-35 NC-40 NC-45 NC-50 NC-55 NC-60 NC-65

135

where, X1 = 1/1 octave band sound pressure level in dB,

X2 = R-values for the throw room using a noise maker and 60 second decay

method,

Minus 10 is for the Standard 70 room correction due to attenuation.

Figure 56. Sound capture interface using the boom microphone. Combined with these calculations was an ambient background sound level, of the current

throw room configuration, to determine the overall NC during each test performed in the

throw room.

136

APPENDIX B

FIELD INSTALLATION TEST RESULTS DIFFUSER DISCHARGE AIRFLOW

ZONE PLOTS

The following figures represent the zone profile for each configuration tested in the

field installations Taguchi mixed array in Chapter 4. The majority of the data resided in

zone three and four as the jet of airflow from the diffuser face spread downward and

away from the diffuser along the ceiling, but the zone two profile existed a short distance

from the diffuser face where throw velocity was the highest. The process used to create

these plots was given in ASHRAE Standard 70-2006. For each field installation

configuration the velocity profiles for each terminal velocity air speed from the array of

draft meters was used to determine throw distance, x, and the corresponding velocity, Vx.

Then calculating the inlet neck area, Ak, and the neck velocity, Vk, the dimensionless

comparison was made below in each figure. These data points were compared to the

similarly plotted zone data points for the Standard 70 baseline data. This comparison led

to the throw ratio values between the field install and the Standard 70 configurations used

in the field installation statistical analysis. Refer to Table 46 for a list of the run numbers

and their corresponding inlet conditions.

137

Table 46. Inlet conditions for field testing runs.

Run #

Diffuser Type Inlet

Duct Height

Type Duct Damper

Damper Type

Approach Angle

1 Square 8 0 Rigid None RndSliding 0 2 Square 10 1.5 Mix Perpendicular RndSliding 23 3 Square 12 3 Flex Parallel RndSliding 45 4 Plaque 8 0 Mix Perpendicular RndSliding 45 5 Plaque 10 1.5 Flex Parallel RndSliding 0 6 Plaque 12 3 Rigid None RndSliding 23 7 Perf Rnd 8 1.5 Rigid Parallel RndSliding 23 8 Perf Rnd 10 3 Mix None RndSliding 45 9 Perf Rnd 12 0 Flex Perpendicular RndSliding 0

10 Mod Core 8 3 Flex Perpendicular SqOpBld 23 11 Mod Core 10 0 Rigid Parallel SqOpBld 45 12 Mod Core 12 1.5 Mix None SqOpBld 0 13 Round 8 1.5 Flex None RndSliding 45 14 Round 10 3 Rigid Perpendicular RndSliding 0 15 Round 12 0 Mix Parallel RndSliding 23 16 Louvered 8 3 Mix Parallel SqOpBld 0 17 Louvered 10 0 Flex None SqOpBld 23 18 Louvered 12 1.5 Rigid Perpendicular SqOpBld 45

Figure 57. Forward throw for Run #1.

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


138

Figure 58. Backward throw for Run #1.


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


139


Figure 61. Right side throw for Run #2.

0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


140



0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


141



0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


142



0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


143


Figure 69. Left side throw for Run #5.

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


144



0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


145



0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


146



0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


147



0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


148



0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


149



0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


150



0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


151



0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


152



0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


153



0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


154



0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


155



0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


156



0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


157

Figure 96. Right side throw of Run #17.

Figure 97. Left side flow for Run #17.

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


158



0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


159

APPENDIX C

CLOSE COUPLING TEST RESULTS DIFFUSER DISCHARGE AIRFLOW ZONE

PLOTS

The following figures are the zone plots for each of the nine configurations in the L9

mixed array used in the Taguchi analysis of the close coupling setup. The majority of the

data was found to be in zone three and four as the jet of airflow from the diffuser face

was directed in a more vertical pattern without a ceiling present, but occasionally a zone

two profile existed a short distance from the diffuser face where throw velocity was

highest, even without a ceiling installed. The process used to create these plots was given

in ASHRAE Standard 70-2006. For each close coupling installation configuration the

velocity profiles for each terminal velocity air speed from the array of draft meters was

used to determine throw distance, x, and the corresponding velocity, Vx. Then calculating

the inlet neck area, Ak, and the neck velocity, Vk, the dimensionless comparison was

made below in each figure. These data points were compared to the similarly plotted zone

data points for the Standard 70 baseline data. This comparison led to the throw ratio

values between the close coupling and the Standard 70 configurations used in the close

coupling installation statistical analysis. Refer to Table 47 for the run numbers and their

corresponding inlet conditions.

160

Table 47. Inlet conditions for close coupling runs.

Run # Diffuser

Type Inlet Duct

Height Damper 1 Square 8 0 None 2 Square 10 1.5 X 3 Square 12 3 Cross 4 Plaque 8 1.5 Cross 5 Plaque 10 3 None 6 Plaque 12 0 X 7 Perf Rnd 8 3 X 8 Perf Rnd 10 0 Cross 9 Perf Rnd 12 1.5 None

Figure 100. Forward throw for Run #1 with no ceiling.

0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


161

Figure 101. Backward throw for Run #1 with no ceiling.

Figure 102. Forward throw for Run #1 with ceiling.

0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


162

Figure 103. Backward throw for Run #1 with ceiling.


0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


163



0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)

Throw/TerminalVelocity 1150 [fpm] Throw/TerminalVelocity 600 [fpm] Throw/TerminalVelocity 300 [fpm] Linear Zone Regression minus 20%

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)

Throw/TerminalVelocity 1150 [fpm] Throw/TerminalVelocity 600 [fpm] Throw/TerminalVelocity 300 [fpm] Linear Zone Regression minus 20%

164



0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


165



0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


166



0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


167



0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


168



0.01

0.1

1

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


169



0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


170


0.01

0.1

1

10

1 10 100

Vx/

Vk

x/Sqrt(Ak)


171

BIBLIOGRAPHY

[1] ASHRAE Handbook of HVAC Applications. Chapter(s): 47.9, 56.1. Atlanta, GA: ASHRAE, 2007.

[2] Hydeman, M., S. Taylor, J. Stein, Taylor Engineering, E. Kolderup, T. Hong, and

Eley Associates. Advanced Variable Air Volume System Design Guide: Design Guidelines. C.E. Commission, Editor. 2003.

[3] ASHRAE Handbook of Fundamentals. Chapter(s): 32.14, 34.1. Atlanta, GA:

ASHRAE, 2009. [4] Landsberger, B., L. Tan, X. Hu. Energy and Acoustic Performance Effects Due to

VAV Duct Design and Installation Practice Variation. Las Vegas, NV: HVAC&R Research, 2008. 14(4).

[5] Koestel, A. and G.L. Tuve. Performance and Evaluation of Room Air Distribution

Systems. ASHRAE Transactions 61:533, 1955. [6] Rock, B.A. and D. Zhu. Designer’s Guide to Ceiling-Based Air Diffusion. Atlanta,

GA: ASHRAE, 2002. [7] Spengler, J. D., Samet, J. M., and McCarthy, J. F. Indoor Air Quality Handbook. New

York, NY: McGraw-Hill Companies, Inc., 2001. [8] Schaffer, M. E. A Practical Guide to Noise and Vibration Control for HVAC Systems.

2nd Edition (pp. 198). Atlanta, GA: ASHRAE, Inc., 2005. [9] ANSI/ASHRAE Standard 70-2006. Method of Testing the Performance of Air

Outlets and Air Inlets. Atlanta, GA: ASHRAE, 2006. [10] Fowlkes, W., and C. Creveling. Engineering Methods for Robust Product Design:

Using Taguchi Method in Technology and Product Development. 1st Edition. Reading, MA: Prentice Hall, 1995.

[11] Reynolds, Douglas. Engineering Principles of Acoustics and Noise Control. Las

Vegas, NV: DDR, Inc., 1997. [12] Goodfellow, H. D., and Tähti, E. Industrial Ventilation Guide Book. San Diego, CA:

Academic Press, 2001. [13] Jackson, Sherri L. Research Methods and Statistics: A Critical Thinking Approach.

3rd Edition (pp. 151-155). Belmont, CA: Wadsworth, 2009. [14] Minitab 16 Statistical Software (2010). [Computer software]. State College, PA:

Minitab, Inc. (www.minitab.com)

172

[15] Landsberger, B., Tan, L., Novosel, D. Underfloor Air Distribution (UFAD). Las Vegas, NV: U.S. Department of Energy, 2006.

173

VITA

Graduate College University of Nevada, Las Vegas

Zaccary A. Poots

Degree:

Bachelor of Science, Mechanical Engineering, 2009 University of Nevada, Las Vegas

Thesis Title: Effects of Inlet Conditions on Diffuser Outlet Performance Thesis Examination Committee:

Chairperson, Douglas Reynolds, Ph. D. Committee Member, Brian Landsberger, Ph. D. Committee Member, Samir Moujaes, Ph. D. Graduate Faculty Representative, Sandra Catlin, Ph. D.

effects of inlet conditions on diffuser outlet performance

Documents